View Full Version : nVidia shader patent (REYES/Raytracing/GI) destined for PS3?
The present invention is related to rendering computer animated video and/or images generally, and to adjusting the origins of rays cast for object-edge positions. The present invention includes identifying a location of a vertex positioned on a perimeter of an object defined in the object scene by a plurality of vertices. The plurality of vertices include the vertex positioned on the perimeter of the object. A shading position that corresponds to, but is offset from, the vertex is then established. A shading value is then computed for the vertex by reference to the shading position. The shading value may be computed, for example, by casting a ray from the shading position.
System and method of adjusting ray origins when shading vertices with rays (http://v3.espacenet.com/textdoc?DB=EPODOC&IDX=US6825840&F=0&QPN=US6825840)
I know I can't help myself, so sue me! :P
But with some talk of what PS3 may do prior to this nVidia announcement, this patent would be a suitable match for PS3 and CELL, no? Especially with Peter Hofstee's recent talk of physics based modelling, ray casting etc. With the importance of shading technology next gen, it's the only patent filed by nVidia that mentions REYES, ray-tracing and global illumination that I could find! :D Even if it's not for PS3, it's still interesting! :)
Guden Oden
08-Dec-2004, 13:31
So what exactly is it that this patent does? It's not immediately obvious for someone that isn't a graphics programmer and versed in patent legalese. :D
Also, what is it that it does different than other raycasting schemes that makes it patentable? :P
london-boy
08-Dec-2004, 13:38
So what exactly is it that this patent does? It's not immediately obvious for someone that isn't a graphics programmer and versed in patent legalese. :D
Also, what is it that it does different than other raycasting schemes that makes it patentable? :P
It has more the's in the patent.
That patent is from Larry Gritz.
Mr. Gritz is an ex Pixar who founded Ex Luna. ExLuna got acquired by NVIDIA a couple of years ago. NVIDIA also settled some patents litigation between Pixar and ExLuna.
IIRC Gritz also developed BMRT (Blue Moon Rendering Tool) the first renderman compliant interface ray tracer (it used some kind of global illumination algorithm too..) who was used as ray server when PRMAN didn't supported ray tracing.
Now he's working on GELATO. That patent is about ExLuna rendering technology, now owned by NVIDIA.
ciao,
Marco
Guden Oden
08-Dec-2004, 14:10
Um, thanks for all the background infos, but like I said in my other post, what does it DO, exactly? :P
Also, I have to ask... Do you have some kind of patent fetish or something? You even keep track of who develops what for whom, and which companies sue each other and then pays the bills... :D
london-boy
08-Dec-2004, 14:13
Um, thanks for all the background infos, but like I said in my other post, what does it DO, exactly? :P
Also, I have to ask... Do you have some kind of patent fetish or something? You even keep track of who develops what for whom, and which companies sue each other and then pays the bills... :D
He's actually a 8 headed, tentacle armed alien, plugged to the internet, and everytime the word "patent" goes through him, he wakes up and posts here.
Guden Oden
08-Dec-2004, 14:27
So why didn't he post when the Abdominal Chruncher was patented, huh? :D
london-boy
08-Dec-2004, 14:30
So why didn't he post when the Abdominal Chruncher was patented, huh? :D
He did, in the fatgeek.com message boards.
That patent is about how efficiently merge REYES algorithm with ray tracing.
Larry Gritz started many years ago (when he worked at Pixar) to study how to improve REYES:
BMRT: A Global Illumination Implementation
of the RenderMan Standard (http://www.siggraph.org/education/materials/HyperGraph/radiosity/jgt96.pdf)
I'm not a patent fetish (OMG :roll:) or an 8 headed alien, I just have a good memory and a nice list of names and terms to check on patent databases. Actually, browsing that list, it takes me 15 minutes (one check per week) to make a full scan of the interesting new patents.. 8)
If this Sony-nVidia collaboration opens up access to these forms of patents/IPs including ones accessible through STI, I can see this being as an extremely bright and fruitfull move. :)
If they had a paticular direction in mind with CELL and PS3, the last thing they would've wanted was to get hindered by patents and make compromises. And this patent is a juicy one to have access to! :P
So they needed Nvidia for REYES? Has Pana anything to do with this? ;)
Fredi
PS: Reading the patent now.
So they needed Nvidia for REYES? Has Pana anything to do with this? ;)
Fredi
PS: Reading the patent now.
Well I hate to speculate, (ahem...yeah right!) but nAo mentioned that ExLuna got bought by nVidia a couple of years ago... and when did nVidia start work on the PS3 GPU...hmm?...Maybe they could implement something from this in hardware for the PS3 GPU? :twisted:
And no, Pana had nothing to do with this... ;)
Wouldn't the fact that Nvidia says they are making a customized version of their next generation gpu rule out REYES unless Nvidia is lying again. :?
For no matter how much better it is it has to haev an Api to stand on and of course devolopers do not like change. Thus microsoft will continue DirectX in it's direction leaving Nvidia out to dry. :cry:
Google, your best friend (or maybe not):
SANTA CLARA, CA — JUNE 13, 2002 — NVIDIA® Corporation (Nasdaq: NVDA), the worldwide leader in visual processing solutions, today announced that the graphics community—including industry luminaries, game developers, middleware vendors, digital content creation (DCC) software vendors and rendering vendors—are overwhelmingly endorsing and adopting the Cg Language and the NVIDIA Cg Toolkit, unveiled earlier today. Industry-leading DCC tool providers, including: Alias|Wavefront, Discreet, Luxology and Softimage, are actively working with NVIDIA to incorporate Cg into their current development plans. The first Cg-based offerings—primarily expected to be software plug-ins—will be available to DCC customers this summer directly within their favorite tools.
...
“Today’s gamer expects the same richness, reality and immersivity in their gameplay as they do with other entertainment media, so the pressure is on the development community to challenge their existing creative boundaries and deliver truly astounding graphics effects. Criterion has always been committed to empowering individual developers with the tools to realize their creative vision, and we fully support NVIDIA’s new Cg Compiler as a positive step towards realizing a real-time REYES/Renderman-style architecture and help narrow the gap between non-real time and real-time graphics.”
Adam Billyard
CTO
Criterion Software
http://developer.nvidia.com/object/IO_20020612_7133.html
That was some months before Sony and Nvidia started working together. Maybe this new GPU is the hardware realistation of the software ideas they had earlier that year and Sony was the perfect partner to realise it.
Fredi
Yes I know but the problem is the programming industry will resist radical change. They always liked to be backwards compatable and REYES wouldn't do that therefore Nvidia maybe hesitant to use it in there next generation of gpu simply because they are afraid if they take the risk no one will follow and they will be hung out to dry. It is not so much wether it is feasible or even better but will they be able to make money when they do it.
Yeah, we need more infos from Sony or Nvidia. ;)
However:
Imagine Stream processors and REYES:
The OpenGL and Reyes rendering pipelines each render complex scenes from similar scene descriptions but differ in their internal pipeline organizations. While the OpenGL organization has dominated hardware architectures over the past twenty years, a Reyes organization differs in several important ways from OpenGL, including a shader coordinate system that supports coherent texture accesses, a single shader in the vertex stage, and tessellation and sampling instead of triangle rasterization. Hardware for the OpenGL pipeline has been well-studied, but the lack of a hardware Reyes implementation has prevented a comparison between the two pipelines. We analyze and compare implementations of an OpenGL and a Reyes pipeline on the Imagine stream processor, a high performance programmable processor for media applications. This comparison both demonstrates the applicability of Reyes for hardware implementation and exposes many issues that architects will face in implementing Reyes in hardware, in particular the need for efficient subdivision algorithms and implementations.
http://shorterlink.com/?1GNQ4F
Fredi
Is it really what Sony after ?
http://film.nvidia.com/page/gelato_gallery.html
Is it really what Sony after ?
http://film.nvidia.com/page/gelato_gallery.html
http://pc.watch.impress.co.jp/docs/2004/0514/kaigai004.jpg
http://pc.watch.impress.co.jp/docs/2004/0514/kaigai005.jpg
They're continually talking of movies and games integration which leads me to believe this...heck this is more plausible than if someone told you Nvidia were doing the PS3's GPU last week! :P
And Hofstee implies this for CELL in the video...
Peter Hofstee's gamespot CELL video (http://hardware.gamespot.com/Story-ST-9398-1266-x-x-x)
Megadrive1988
09-Dec-2004, 00:24
I watched the Hofstee video. extremely interesting to see someone who is at the heart of Cell's design talk about it. I could easily understand what he was saying. thanks for posting the vid.
even if PS3 does not reach 1,000 times the performance of PS2 as Sony stated in fall 1999, at least they might reach the '18,000 times PS1' performance, which was a mid-to-late 1990s desire of game developers, according to Sony.
i.e. Sony states that PS2 is 300 times the performance of PS1. therefore PS3 would only need to be 60 times the performance of PS2. unless my math is flawed :) again, going by Sony's figures and Sony's thinking.
They always liked to be backwards compatable and REYES wouldn't do that therefore Nvidia maybe hesitant to use it
Except maybe in a proprietary machine that doesn't need to be backwards compatible with every hardware accelerated game on the x86 platform?.. something like, a game console? =P
Sure, they couldn't slap the tech into thier next generation PC gpu, but how many PS2s are out there again? ;)
They always liked to be backwards compatable and REYES wouldn't do that therefore Nvidia maybe hesitant to use it
Except maybe in a proprietary machine that doesn't need to be backwards compatible with every hardware accelerated game on the x86 platform?.. something like, a game console? =P
Sure, they couldn't slap the tech into thier next generation PC gpu, but how many PS2s are out there again? ;)
Would they really want to sacrifice their PC product line even worse than what they did with geforce fx? I suppose they could take a break for a year from pc graphics, but if PC graphics remain alive and well then ati will be the one releasing a new product in the fall and nvidia releasing one in spring of next year that can just barely compete with the previous product, let alone the new one ati is releasing in the spring.
randycat99
09-Dec-2004, 04:22
Looking at a more grand, ahem, vision ;), perhaps nVidia is orchestrating a position that looks farther than just what to sell next year. For a long time, it has been premised (in the developer community) when could there be a logical break from the nasty x86 legacy once and for all? It stifles processor design, complicates software development, and more than likely has slowed the advance of high-performance CG. Yet, year after year it has to be supported because it is that entrenched into the industry. So is it possible that nVidia has seen the writing on the wall (or maybe even actively drawn the lines in the sand) to jump in Sony's boat to make the "impossible", possible? They want to not only take first divs on a bold new computing paradigm, but they want to take an active role in molding it using their well-honed graphics technology expertise. This new computing paradigm will be something crafted from the synergies of processors, OS, and graphics functionality designed from scratch and out of the knowledge of how it should be done, given a clean sheet of paper. No longer is there a subservience to legacy x86 or existing hardware born to survive in the x86 world.
So you can premise that they will miss the hardware cycle for PC GPU's next year, but do not forget that they will be well revenued for years to come for the work they have put into PS3. On top of that, they will be well out of the gate and running in this exciting new computing paradigm (if it really exists, and if it indeed proves viable). They will be deep into the development curve for creating successive videocard products for the computers that are born out of this new computing paradigm.
Now I'm not saying that this is for certain. I'm not even saying it will succeed if there is indeed a "grand plan". It's just a wacky theory. However, I believe it sheds light on how nVidia has positioned themself such that it "seems" they are abandoning territory that would conventionally be seen as their home turf. For lack of less provocative terms, the real issue to be observed could be that they are "abandoning a sinking ship", rather than an old, trusted money-making territory.
I don't think the x86 pc market will die unless 1 of two things happens...
Microsoft decides it should die.
Sony includes computer like functionality that is good enough for the average person in their Ps3.
randycat99
09-Dec-2004, 05:09
Of course it will die some day. :P Everything does...
More seriously, this might be something that self-implodes or just dies a slow death (by no small measure of MS spending its very last dollar to prevent it dying trend).
Also, don't count out the possibility of non-x86-based Sony desktops to appear somewhere down the line. If it can fulfill the basic needs of the typical home user while shedding the Windows-related downsides of current desktops, that could just turn out to be a big hit. ...and if the Cell-based internals manage to deliver an undismissable wallop in graphics and multimedia functionality, that just may be enough to seal the fate of the x86 era. True, there is a lot hanging in the balance for that to happen, but it certainly is within the realm of possibility.
This was posted from Nirey, a respected and trusted poster from PSINEXT,
Hey sorry guys, i have not posted in so long since last time. But anyways i am sorry i couldn't say anything about this sweet deal until it was announced since i was under very tight NDA's. Anyways, now it's official, and i would like to say that this is a co-joint collaboration between sony in-house graphics technology and Nvidia technology. What this means is that this "custom-gpu" is a totally different architecture that is not based on an existing architecture from Nvidia.
More will come soon.
http://psinext.com/forums/viewtopic.php?p=84712#84712
Hmmm...curioser and curioser...REYES?...Raytracing?...GI? :twisted:
More will come soon.
Soon can't come soon enough :!:
More will come soon.
Soon can't come soon enough :!:
Indeed...the ideal time to introduce something like this would be in a clean-sheet, legacy free console architecture guaranteed tens of millions in sale, even with a less than perfect introduction! :idea: 8)
For a limited time, NVIDIA is making Gelato 1.1 available at approximately 45% off the regular price:
Software Only
$1,799 (includes Gelato 1.1 and 1 year M&S), save $1,476
Gelato/Quadro FX 700 Bundle
$2,099 (includes Gelato 1.1, 1 year M&S, and Quadro FX 700), save $1,691
Gelato/Quadro FX 3000 Bundle
$2,499 (includes Gelato 1.1, 1 year M&S, and Quadro FX 3000), save $2,076
For more information on the Quadro FX boards, see http://film.nvidia.com/page/quadro.html
Contact Matt Jefferson, NVIDIA Digital Film Group Sales, at gelatosales@nvidia.com or (951) 898-8863 for details on how to order Gelato at this special price.
Special prices good through 14 January 2005. Get it before it melts.
http://film.nvidia.com/page/gelato.html
Hmmm...there seems to be a huge nVidia sale on with it's current Gelato software/hardware packages...hmmm...I wonder why! :?: ;)
PC-Engine
09-Dec-2004, 13:54
If Apple with it's OSX and flashy computers can't make a significant dent in the Windows/Intel/AMD market, how will SONY and the unproven CELL kill anything? Unless they start dumping hardware and software, I don't see it happening.
If Apple with it's OSX and flashy computers can't make a significant dent in the Windows/Intel/AMD market, how will SONY and the unproven CELL kill anything? Unless they start dumping hardware and software, I don't see it happening.
Well, maybe it will have something OSX is lacking, how about a steady supply of kick ass, triple A, install hassle free games?
PC-Engine
09-Dec-2004, 14:22
If Apple with it's OSX and flashy computers can't make a significant dent in the Windows/Intel/AMD market, how will SONY and the unproven CELL kill anything? Unless they start dumping hardware and software, I don't see it happening.
Well, maybe it will have something OSX is lacking, how about a steady supply of kick ass, triple A, install hassle free games?
Why buy a computer for that when you can get it from a console? :lol:
Why buy a computer for that when you can get it from a console?
Why buy a computer and a console? :lol:
From now on do not make a new thread about this . This about the 18th thread we have had in the last month about reyes .
PIck one thread and keep it all in there .
PC-Engine
10-Dec-2004, 12:00
Why buy a computer for that when you can get it from a console?
Why buy a computer and a console? :lol:
Because a computer can do so much more than a game console unless you think PS3 disguised as a computer will be replacing PCs and Macs. :lol:
Why buy a computer for that when you can get it from a console?
Why buy a computer and a console? :lol:
Because a computer can do so much more than a game console unless you think PS3 disguised as a computer will be replacing PCs and Macs.
Well, who knows. We got some talk about cell-based workstations, Sony gave Linux to PS2 users, Sony got a very appealing line of well designed VAIO computers ... let's throw in some ifs: PC designed around the PS3 platform, HDD, some OS (e.g. Linux based), able to run signed PS3 game code, able to run a bunch of Linux apps (e.g. Firefox, OpenOffice, Thunderbird, mplayer, gimp), some effort to make it usuable for end-users (flashy GUI, some easy to use package installation system), utilizing the advantages of the cell architecture (networkbased computational loadsharing, no computational x86/ppc islands anymore)... it perhaps would work, in a kinda Amiga/ST way. Sure it wouldn't kill x86 overnight, but it would be an option and it would be something to build on. I would like it, even though it's unlikely to happen *sigh*
... and wasn't there a rumor, that MS was going to release three XBox2-based platforms, one being some kind of PC setup?
rabidrabbit
10-Dec-2004, 14:12
A "computer" for home use does not need to perform all the functions, run similar types of software as a "computer" for business use.
A mainstream "computer" user doesn't need to run cad-applications, heavy database software, complicated spreadsheet, publishing software...
A home "computer" user is not the same as a geek computer user.
A home "computer" user wants to edit/view/listen photographs, video, music, maybe write a few sheets of text, e-mail, internet, video/voice/text chat, play games.... mostly entertainment. Some "heavy" user might want to keep track of family income, bills etc...
How many (non geek) PC users are really using their computers for other than entertainment.
london-boy
10-Dec-2004, 14:21
A "computer" for home use does not need to perform all the functions, run similar types of software as a "computer" for business use.
A mainstream "computer" user doesn't need to run cad-applications, heavy database software, complicated spreadsheet, publishing software...
A home "computer" user is not the same as a geek computer user.
A home "computer" user wants to edit/view/listen photographs, video, music, maybe write a few sheets of text, e-mail, internet, video/voice/text chat, play games.... mostly entertainment. Some "heavy" user might want to keep track of family income, bills etc...
How many (non geek) PC users are really using their computers for other than entertainment.
Mmmm I don't think it's up to Sony to decide what people can do with their own computers.
Besides, what i was thinking was, a PS3 will in the end cost as much as a high-end PC graphics card, around 300 quid.
So what's stopping Sony to include at least part of the PS3 architecture, up to them which part, to their Vaio line, in order to accellerate specific tasks.
Obviously they will have to come up with new motherboards, and it might incurr problems with Intel or AMD and whatever vendors they usually put in their Vaio line of computers...
But they could easily "hide" Cell in there somewhere, if they wanted to.
PC-Engine
10-Dec-2004, 14:37
A "computer" for home use does not need to perform all the functions, run similar types of software as a "computer" for business use.
A mainstream "computer" user doesn't need to run cad-applications, heavy database software, complicated spreadsheet, publishing software...
A home "computer" user is not the same as a geek computer user.
A home "computer" user wants to edit/view/listen photographs, video, music, maybe write a few sheets of text, e-mail, internet, video/voice/text chat, play games.... mostly entertainment. Some "heavy" user might want to keep track of family income, bills etc...
How many (non geek) PC users are really using their computers for other than entertainment.
People want the option of having many differing levels of computers from power users all the way down to light users. Apple already does many of these things yet they still can't increase their marketshare beyond niche status. How is a light PS3 computer gonna even compete beyond even Apples tiny share? PCs are dirt cheap nowadays so cost isn't even going to be an advantage for a so called PS3 computer.
rabidrabbit
10-Dec-2004, 14:40
Do they really? Or is it just because they really have no other option?
I don't think Sony is even targeting the "power users".
and it's arguable how many of the so called "power users" really are power users. They might like themselves to be called "power users" and they might think they are "power users", but in reality they use their PC's very "lightly" (updating your drivers monthly, re-installing the OS every year, trying to get things to work... that doesn't necessarily a "power user" make :wink: )
PC-Engine
10-Dec-2004, 14:44
Do they really? Or is it just because they really have no other option?
Yes they do really. If they didn't then every PC being sold would be a light version. The more powerful PCs cost more than the light versions. People wouldn't buy the higher priced ones if they didn't want the higher performance. It doesn't matter what the power users actually use that power for if they indeed even use it. The point is they want choices since it's their money. It's like cars. Do you have to actually race them for you to go out and buy a Porsche? Most sportscars never touch a race track.
london-boy
10-Dec-2004, 14:45
A "computer" for home use does not need to perform all the functions, run similar types of software as a "computer" for business use.
A mainstream "computer" user doesn't need to run cad-applications, heavy database software, complicated spreadsheet, publishing software...
A home "computer" user is not the same as a geek computer user.
A home "computer" user wants to edit/view/listen photographs, video, music, maybe write a few sheets of text, e-mail, internet, video/voice/text chat, play games.... mostly entertainment. Some "heavy" user might want to keep track of family income, bills etc...
How many (non geek) PC users are really using their computers for other than entertainment.
People want the option of having many differing levels of computers from power users all the way down to light users. Apple already does many of these things yet they still can't increase their marketshare beyond niche status. How is a light PS3 computer gonna even compete beyond even Apples tiny share? PCs are dirt cheap nowadays so cost isn't even going to be an advantage for a so called PS3 computer.
Why not? PS3 itself will cost much less than a PC, and it will cost much less than a PC that outputs the same level of graphics, at least for the first year or so.
rabidrabbit
10-Dec-2004, 14:46
Do they really? Or is it just because they really have no other option?
Yes they do really. If they didn't then every PC being sold would be a light version. The more powerful PCs cost more than the light versions. People wouldn't buy the higher priced ones if they didn't want the higher performance. It doesn't matter what the power users actually use that power for if they indeed even use it. The point is they want choices since it's their money. It's like cars. Do you have to actually race them for you to go out and buy a Porsche? Most sportscars never touch a race track.
You can't play Half-Life 2 on a "light version PC" :lol:
What do you think drives the sales of those high end PC´s, that people need all that power to run huge databases in Access?
Edited to include the quotes
PC-Engine
10-Dec-2004, 14:49
You can't play Half-Life 2 on a "light version PC"
What do you think drives the sales of those high end PC´s, that people need all that power to rin huge databases in Access?
And I never said you could, since I never claimed that higher power PCs were mainly used for games. You're confusing the issue anyway. You need a powerful video card AND a powerful cpu to play highend games. A powerful PC doesn't necessarily mean a powerful videocard. Most people who buy powerful PCs want it just because of extra piece of mind not because they use it for playing games like HL2. :lol:
london-boy
10-Dec-2004, 14:51
Do they really? Or is it just because they really have no other option?
Yes they do really. If they didn't then every PC being sold would be a light version. The more powerful PCs cost more than the light versions. People wouldn't buy the higher priced ones if they didn't want the higher performance. It doesn't matter what the power users actually use that power for if they indeed even use it. The point is they want choices since it's their money. It's like cars. Do you have to actually race them for you to go out and buy a Porsche? Most sportscars never touch a race track.
Doesn't matter, that's a flawed argument, high end video cards (which will cost MORE than a whole PS3) are bought to play the latest games with full eye candy on.
People don't buy 6800U's to run CAD or Excel.
rabidrabbit
10-Dec-2004, 14:52
rabbit, I've added a few more comments to my previous post.
ok, but I'm too lazy to turn the page :lol:
PC-Engine
10-Dec-2004, 15:02
Doesn't matter, that's a flawed argument, high end video cards (which will cost MORE than a whole PS3) are bought to play the latest games with full eye candy on.
People don't buy 6800U's to run CAD or Excel.
I'm full aware that a highend graphics card would cost more than a whole PS3. That's only relevent if you're talking about the hardcore PC gamer which is a small group relative to the non hardcore and non gamer. The majority of PCs being sold is not for game playing. It's being used for other things. Just because it's being used for other light stuff, doesn't mean people don't want a fast cpu, RAM, DVD burner drive etc. Besides new videocards come out every few months anyway so the PS3 PC would already be surpased by the time it makes it onto shelves and those same PS3s graphics won't be able to be updated until PS4 comes out. :lol:
rabidrabbit
10-Dec-2004, 15:05
Majority of PCs sold are to companies, I think ;)
But the majority of home PC's sold need not be as high end as they are.
Many people buying them are easy to sell a high end PC to, 'cos they know little what they really need or want.
PC-Engine
10-Dec-2004, 15:10
Majority of PCs sold are to companies, I think ;)
But the majority of home PC's sold need not be as high end as they are.
Many people buying them are easy to sell a high end PC to, 'cos they know little what they really need or want.
Yes but that's a given not to mention companies don't buy them for games. :lol:
Sure but people don't want SONY to tell them what they need and want and how to spend their own money. :wink:
rabidrabbit
10-Dec-2004, 15:11
Majority of PCs sold are to companies, I think ;)
But the majority of home PC's sold need not be as high end as they are.
Many people buying them are easy to sell a high end PC to, 'cos they know little what they really need or want.
Yes but that's a given not to mention companies don't buy them for games. :lol:
Sure but people don't want SONY to tell them what they need and want and how to spend their own money. :wink:
No, but do they need the salesman from your local electronics shop?
PC-Engine
10-Dec-2004, 15:15
Majority of PCs sold are to companies, I think ;)
But the majority of home PC's sold need not be as high end as they are.
Many people buying them are easy to sell a high end PC to, 'cos they know little what they really need or want.
Yes but that's a given not to mention companies don't buy them for games. :lol:
Sure but people don't want SONY to tell them what they need and want and how to spend their own money. :wink:
No, but do they need the salesman from your local electronics shop?
You're not required to ask for sales advice. If you already know what you want you just buy it no sales advice required.
london-boy
10-Dec-2004, 15:24
I'm full aware that a highend graphics card would cost more than a whole PS3. That's only relevent if you're talking about the hardcore PC gamer which is a small group relative to the non hardcore and non gamer. The majority of PCs being sold is not for game playing. It's being used for other things. Just because it's being used for other light stuff, doesn't mean people don't want a fast cpu, RAM, DVD burner drive etc. Besides new videocards come out every few months anyway so the PS3 PC would already be surpased by the time it makes it onto shelves and those same PS3s graphics won't be able to be updated until PS4 comes out. :lol:
But that's my point: we ARE talking about the hardcore PC gamers which are a small group relative to the non hardcore and non gamer, because only the hardcore gamer buys a 6800U or X800PE, cause he wants the maximum detail available to him at the highest framerates (freak).
PS3 will, for a limited time only, produce graphics that are much superior to those available to the geekiest highest end hardestcore gamer, and at a much lower price.
I agree, and have said, that Sony shouldn't be the one to decide what people want to do with their PCs, nor should anyone else.
How is a light PS3 computer gonna even compete beyond even Apples tiny share? PCs are dirt cheap nowadays so cost isn't even going to be an advantage for a so called PS3 computer.
Why would they have to? If PS3 is equally successful as PS is and PS2 will be, you can expect a 100 million userbase towards the end of its lifespan. If the PS3 can already take care of most basic functions the average consumer buys a home PC for and one is already sitting there - why buy a more expensive PC? One of the general fears in buying PCs today is its complicated nature, required upgrades and it being out-dated soon. PCs are already on a decline by a certain audience - and it's only pushed by again a different crowd that buys it primarely as a gaming platform.
It's pretty easy to see where it can all lead to. Why again do you think Microsoft got into the industry and why do we read about "fight over the living room"? If Microsoft was so safe about its own future in the PC realm, why would they participate in being in the living room? Answer this and you can pretty much conclude/answer your own questions brought forward above.
PC-Engine
11-Dec-2004, 08:06
That's the thing, nobody here can answer that question.
randycat99
11-Dec-2004, 08:37
The answer will come as a desktop platform that manages to deliver so much more in capability than any bargain basement emachine could ever dream of, while coming in cheaper than the "built to the hilt" Wintel box, runs a wide range of software (perhaps even under multiple OS's, as well), while simultaneously circumventing all of the common Windows foibles (MS code bloat, registry rot, just plain inexplicably wierd behavior, and legions of security exploits that leave the platform leaking like a sieve). It's a tall order, but not an implausible order coming from a platform born from a clean slate, rather than waiting for Windows to "evolve" to the highly capable, yet secure, OS it so desperately promises to be with every new "release".
PC-Engine
11-Dec-2004, 09:13
People like the option of being able to upgrade individual parts of their computers whenever they want. That will be difficult for a closed box PS3 computer. If they decide to make it an open box using industry standard interfaces, the cost will go up. Adding a simple HDD will already raise the cost by at least $50.
randycat99
11-Dec-2004, 09:31
The "people" you mention are a niche group, as you were so careful to describe earlier on. It will be irrelevant to the widespread adoption of a successor platform. ...and since when would adding a HD be any more difficult than in a Wintel box?! As has already been implied, CPU and graphics performance will already be in excess in what the typical user (or even the power user, for that matter) will need. So no great need to pick and choose and upgrade there. However, since you bring it up, "upgrading" may well become an obsolete pasttime, in exchange for scaleability. Need more resources? Your nearby PS3 may well double your capability by just connecting it to your Cell desktop with a high-speed link. Need more? Daisychain another PS3 to the setup. Falling behind after 5 yrs or so? Replace both units with a PS4 and connect that to your Cell desktop.
This all sounds fanciful, but not at all impossible once x86 is left behind and a clean-sheet design with implicit scaleability native to its architecture takes hold. Given that kind of potential and that sort of userbase, what developer wouldn't try this new boat?
PC-Engine
11-Dec-2004, 10:05
I guess it would be a neat idea for the average person to only have to daisy chain systems together for more power. If they can implement that with a single unified power supply and keep the units small, cool, and quiet, it would be an attractive option for the average person and even some of the hardcore. Of course if the software (games mostly) is designed to automatically scale with the additional nodes of computing power, it would be even better. Howver the units would have to be pretty cheap after a year or two for the average person to buy more than one just for the sake of boosting performance.
pahcman
11-Dec-2004, 11:09
rather than waiting for Windows to "evolve" to the highly capable, yet secure, OS it so desperately promises to be with every new "release".
Why wait? Windows XP.
You think a new born PS OS can match XP breadth of functions? Hence one of the reasons why consoles have yet replace PC generations after generations, predictions after predictions. Not to forget the cost ceiling too.
rather than waiting for Windows to "evolve" to the highly capable, yet secure, OS it so desperately promises to be with every new "release".
Why wait? Windows XP.
You think a new born PS OS can match XP breadth of functions? Hence one of the reasons why consoles have yet replace PC generations after generations, predictions after predictions. Not to forget the cost ceiling too.
It's sometimes easy in a thread to loose sight of the woods from the trees...
It's not about any particular OS or hardware or a particular company...It's about a clean sheet hardware/software design and patents/IPs being infringed. And the particular IP in question is the topic of this thread, this nVidia shader patent and what this could bring to the table for Sony and PS3.
As I posted earlier,
If this Sony-nVidia collaboration opens up access to these forms of patents/IPs including ones accessible through STI, I can see this being as an extremely bright and fruitfull move. :)
If they had a paticular direction in mind with CELL and PS3, the last thing they would've wanted was to get hindered by patents and make compromises. And this patent is a juicy one to have access to! :P
and
...
Indeed...the ideal time to introduce something like this would be in a clean-sheet, legacy free console architecture guaranteed tens of millions in sale, even with a less than perfect introduction! :idea: 8)
So it's not about Windows or Linux or Mac OSX they just happen to facilitate the IPs. Whatever this CELL OS will be is irrelevant but the mechanics at work to bring something like this to fruitition has chosen this particular allegiance and path. And MS are aware of this battle in the living room through a console as a trojan horse. Hence rumours of PC like multiple versions of Xenons and bridging PC gaming via XNA. ;)
I have a question for all of you.
Is true that the actual GPU/VPU cannot do Raytracing in a good speed because Raytracing is a trigometric calculations technique and the VPU/GPU are specialized in Vectorial calculations technique?
rendezvous
11-Dec-2004, 13:53
I have a question for all of you.
Is true that the actual GPU/VPU cannot do Raytracing in a good speed because Raytracing is a trigometric calculations technique and the VPU/GPU are specialized in Vectorial calculations technique?
No, raytracing is all about vectors.
Is true that the actual GPU/VPU cannot do Raytracing in a good speed because Raytracing is a trigometric calculations technique and the VPU/GPU are specialized in Vectorial calculations technique?
that's completely false, moreover trigonometric calculations and vectorial calculations can be deeply related.
rather than waiting for Windows to "evolve" to the highly capable, yet secure, OS it so desperately promises to be with every new "release".
Why wait? Windows XP.
You think a new born PS OS can match XP breadth of functions? Hence one of the reasons why consoles have yet replace PC generations after generations, predictions after predictions. Not to forget the cost ceiling too.
Why need an OS? Have each application completely stand alone so its quality is not tied to the OS.
london-boy
11-Dec-2004, 17:16
rather than waiting for Windows to "evolve" to the highly capable, yet secure, OS it so desperately promises to be with every new "release".
Why wait? Windows XP.
You think a new born PS OS can match XP breadth of functions? Hence one of the reasons why consoles have yet replace PC generations after generations, predictions after predictions. Not to forget the cost ceiling too.
Why need an OS? Have each application completely stand alone so its quality is not tied to the OS.
:shock:
That would need a single common architecture. Or else each developer would have to program their application for each and every different configuration of PC there is out there, and there are LOTS of different configs...
Also, how would you manage all your programs?
There has to be an OS, even a very simple one, but there needs to be one.
I'm not exactly sure why Windows is now as bloated as it is, how it happened and why it takes something like 10-20GB although coming from a simple 650MB CD, but no OS at all aint gonna work.
Now even mobile phones have OSs, because it's easier for the end user to manage a device if there is a program that takes care of all the stuff "in the background", and OS.
rather than waiting for Windows to "evolve" to the highly capable, yet secure, OS it so desperately promises to be with every new "release".
Why wait? Windows XP.
You think a new born PS OS can match XP breadth of functions? Hence one of the reasons why consoles have yet replace PC generations after generations, predictions after predictions. Not to forget the cost ceiling too.
Why need an OS? Have each application completely stand alone so its quality is not tied to the OS.
:shock:
That would need a single common architecture. Or else each developer would have to program their application for each and every different configuration of PC there is out there, and there are LOTS of different configs...
Also, how would you manage all your programs?
There has to be an OS, even a very simple one, but there needs to be one.
I'm not exactly sure why Windows is now as bloated as it is, how it happened and why it takes something like 10-20GB although coming from a simple 650MB CD, but no OS at all aint gonna work.
Now even mobile phones have OSs, because it's easier for the end user to manage a device if there is a program that takes care of all the stuff "in the background", and OS.
Single common architecture = playstation 3.
And maybe it would make you put in a disk when you want to use the application, or they could have a simple menu from where you could select the application you want to use.
london-boy
11-Dec-2004, 17:39
Single common architecture = playstation 3.
And maybe it would make you put in a disk when you want to use the application, or they could have a simple menu from where you could select the application you want to use.
But that's not a PC then, that's still a console. I might want to use more than one application at once. And also it would be nice not to have to insert a disc everytime i need the application to start.
An OS will always be needed. Shame that we only have Windows now, and it doesn't give people much of a choice or even a nice idea of OS's in general.
Single common architecture = playstation 3.
And maybe it would make you put in a disk when you want to use the application, or they could have a simple menu from where you could select the application you want to use.
But that's not a PC then, that's still a console. I might want to use more than one application at once. And also it would be nice not to have to insert a disc everytime i need the application to start.
An OS will always be needed. Shame that we only have Windows now, and it doesn't give people much of a choice or even a nice idea of OS's in general.
I have no idea if you are anti-pc or just ignorant of such but there are more OS's than just Windows. Also NO Windows installation ever reached around 10 or 20 gig or are you one of those people who thing that everything that is installed is the OS.
london-boy
11-Dec-2004, 18:25
I have no idea if you are anti-pc or just ignorant of such but there are more OS's than just Windows. Also NO Windows installation ever reached around 10 or 20 gig or are you one of those people who thing that everything that is installed is the OS.
:| No whay!!!??? Their r moer OSs ohter then Windoz!!11!!????? And u meen taht My Docuemnts aint included in the figurez??? :roll:
Obviously my post wasn't clear enough. I apologise. :roll:
Fafalada
11-Dec-2004, 20:22
Why wait? Windows XP.
You think a new born PS OS can match XP breadth of functions?
Given that half of that "breadth of functions" is crap that makes the system worse then it could have been, I don't see why not.
Now if you said Windows 2000, I would be more inclined to agree with you.
Of course arguing about OS functionality is all fluff - what defines a successfull OS is the software support (not unlike consoles :P) and that's where a new born OS will obviously always have an uphill battle.
In re-reading this (http://graphics.cs.uni-sb.de/~jofis/SaarCOR/DynRT/Schmittler_et_al-Realtime_Ray_Tracing_of_Dynamic_Scenes_on_an_FPGA_ Chip.pdf) paper on ray tracing recently:
They mention that they can do real basic ray tracing with an FPGA they programmed running at 4 Gflops. However, they also mention:
Probably the important missing feature is programmable shading, which would finally provide the full capabilities of ray tracing to applications and users. Together with the ability to easily implement global effects this seems like the killer application for ray tracing.
Wouldn't the CELL + nVidia ROP provide the essential combination for this to happen?
Could this be why they seemingly won't use nVidia for the whole graphics pipeline so that they can leave this road open?
Wouldn't the CELL + nVidia ROP provide the essential combination for this to happen?
APUs local memory is so small that one can't just pretend to run a plain RT implementation on them without issuing a lot of external memory references via DMA requests. That's the kind of things one doesn't want to happen on a stream processor.
Maybe an advanved RT implementation on CELL could exploit rays coherence thus grouping queries on some spatial subdivision structure to a void a lot of scattered memory read operations, otherwise A|SPUs built-in dma prefetch mechanism would be uneffective!
Once DeanoC wrote that a future PS4 and XBOX3 comparison would be based on external memory latencies and not raw flop/s figures..I couldn't agree more ;)
One key performance figure I would like to know about A|SPUs is the time needed to switch thread and it would be nice to have some kind of support for fast registers saving and restoring, or fast registers banks switching.
A|SPUs flexibility is a given..one can do everything, the real problem is how to do everything keeping A|SPUs don't run idle most of the time.. ;)
Could this be why they seemingly won't use nVidia for the whole graphics pipeline so that they can leave this road open?
Why are you saying they seemingly won't use nvidia IP for the whole pipeline? I don't think we have any data about or am I missing something here?
Your question is not unexpcted anyway, at some point GPU manufacters would have to introduce some very fast and flexible calculations core...maybe something along the lines of CELL architecture could be one of the answers and NVIDIA would like to use it.
Too bad I think that's not the case..I'm still skeptic about CELL stuff being into PS3 GPU ;)
ciao,
Marco
Since CELL is a stream processor, here's an interesting ppt on how to do raytracing on a stream processor by mapping to a GPUs programmable fragment processor...
Ray Tracing on Programmable Graphics Hardware ppt. (http://online.cs.nps.navy.mil/DistanceEducation/online.siggraph.org/2002/Papers/13_GraphicsHardware/purcell.ppt)
So could the NV GPU just be one big fragment processor complementing the CELLs programmable vertex processor :?:
Since CELL is a stream processor, here's an interesting ppt on how to do raytracing on a stream processor by mapping to a GPUs programmable fragment processor...
Ray Tracing on Programmable Graphics Hardware ppt. (http://online.cs.nps.navy.mil/DistanceEducation/online.siggraph.org/2002/Papers/13_GraphicsHardware/purcell.ppt)
So could the NV GPU just be one big fragment processor complementing the CELLs programmable vertex processor :?:
Are you sure about the link?
I only get "The page cannot be displayed".
Since CELL is a stream processor, here's an interesting ppt on how to do raytracing on a stream processor by mapping to a GPUs programmable fragment processor...
Ray Tracing on Programmable Graphics Hardware ppt. (http://online.cs.nps.navy.mil/DistanceEducation/online.siggraph.org/2002/Papers/13_GraphicsHardware/purcell.ppt)
So could the NV GPU just be one big fragment processor complementing the CELLs programmable vertex processor :?:
Are you sure about the link?
I only get "The page cannot be displayed".
http://online.cs.nps.navy.mil/DistanceEducation/online.siggraph.org/2002/Papers/13_GraphicsHardware/purcell.ppt
It works for me...it's a direct link...PM me if it still doesn't work...
Could it be that next gen differentiator be that Cell architeture would allow ray tracing/Reyes and Xenon next system a more traditional approach to shadeing yet with much more performance than todays gpu, per pixel lighting, etc...?
Wouldn't the CELL + nVidia ROP provide the essential combination for this to happen?
APUs local memory is so small that one can't just pretend to run a plain RT implementation on them without issuing a lot of external memory references via DMA requests. That's the kind of things one doesn't want to happen on a stream processor.
Maybe an advanved RT implementation on CELL could exploit rays coherence thus grouping queries on some spatial subdivision structure to a void a lot of scattered memory read operations, otherwise A|SPUs built-in dma prefetch mechanism would be uneffective!
Yes you are right in that the APU memory is limited. But I am not ready to give up yet. Here are some reasons supporting the potential for a ray tracing form of rendering with a CELL + nVidia architecture:
1) As Jaws mentioned, GPU ray tracers have been around now as a proof of concept of ray tracing on a streaming processor. They also highlight just how wonky it is to use GPUs as ray tracing engines. William Dally talking about the Imagine Stream Processor from Stanford also explicitely mentioned ray tracing in a streaming architecture. I agree it is not that great a match for RT but it is a potential avenue.
2) There are opportunities for coherence in rays especially for rays with common origins (shooting multiple shadow rays at a point to multiple lights) or rays with common directions.
3) Hofstee in his video mentions global illumination (as I recall)
4) The paper I linked to above also includes some discussion about memory bandwidth and access patterns. They indicate that ray tracing does require pretty random access patterns but that this keep different banks of memory utilized fully instead of everyone clobbering one chunk.
Their own ray tracing pipeline requires 75 KB of memory per pipeline which could fit in an APU memory plus an overall cache to prevent triangles from being intersected multiple times. They have 64 threads going on at a time. They estimate bandwidth (worst case) of 2GB/s bandwidth for 1024x768x60Hz which is easy for Cell.
They mention that their 90Mhz FPGA is equivalent to an 8-12Ghz CPU and is about 3-5x GPU implementation because of its special purpose design. The Cell is not this special purpose but does have part of the streaming advantage along with the clockspeed advantage. As you keenly mentioned, the critical element is the special bits they added for keeping track of traced objects and rebuilding ray tracing object hierarchies as objects move. These kinds of things that are special purpose to their chip and which could not be efficiently ported over are the unknowns. If these unkowns are critical then yes the dream is over (for now).
Ray tracing is not sooo bad because it is completely demand driven. You are only processing pixels/triangles that contribute to the final scene and in that sense you are only asking of the memory what is relevant.
Once DeanoC wrote that a future PS4 and XBOX3 comparison would be based on external memory latencies and not raw flop/s figures..I couldn't agree more ;)
One key performance figure I would like to know about A|SPUs is the time needed to switch thread and it would be nice to have some kind of support for fast registers saving and restoring, or fast registers banks switching.
A|SPUs flexibility is a given..one can do everything, the real problem is how to do everything keeping A|SPUs don't run idle most of the time.. ;)
Yes I think if you look at the supercomputer market you can see that Cray has made advances not just because of pure speed but because their data transfer systems in the machine are low latency (the lowest around).
I think also that Longhorn has some specific requirements about rendering latency and handling huge numbers of threads (1000s). I believe they want every icon to be in 3D and in a separate thread. Stuff like that must cause nVidia and ATI a lot of headaches.
Could this be why they seemingly won't use nVidia for the whole graphics pipeline so that they can leave this road open?
Why are you saying they seemingly won't use nvidia IP for the whole pipeline? I don't think we have any data about or am I missing something here?
Your question is not unexpcted anyway, at some point GPU manufacters would have to introduce some very fast and flexible calculations core...maybe something along the lines of CELL architecture could be one of the answers and NVIDIA would like to use it.
Too bad I think that's not the case..I'm still skeptic about CELL stuff being into PS3 GPU ;)
ciao,
Marco
I am just taking a hint from the patent which shows the Cell top side and Pixel Engine bottom side as well as an image from a patent (not sure which one) which shows the APUs set up as a vertex processing pipeline. Not an airtight case by any means but it's all we have so far.
http://online.cs.nps.navy.mil/DistanceEducation/online.siggraph.org/2002/Papers/13_GraphicsHardware/purcell.ppt
It works for me...it's a direct link...PM me if it still doesn't work...
I got it here if anyone wants
http://cggmwww.csie.nctu.edu.tw/seminar/GPU/2003.11.06.jhlu.ray.tracing.on.programmable.graphi cs.hardware.ppt
Interesting R8500 is better than GF3, I am right?
Raytracing on GPUs seem like a dead end to me.
The benefit of raytracing is in the way it allows a more correct handling of refraction and specular reflection. This makes it ill suited for a stream based approach because it introduce a data dependent step that has, potentially, highly irregular access patterns (cache busting).
I can see GPUs as accelerating first intersection rays (which is also what the above paper deals with), but these account for just a fraction of the work a heavy duty raytracer has to do.
Cheers
Gubbi
How many possibilities are that NVidia plans one GPU for PS3 with one 'shader Core'(vertex+pixel)?
For example 48 ALU in 2x24 pipeline or 3x16 pipeline.
How many possibilities are that NVidia plans one GPU for PS3 with one 'shader Core'(vertex+pixel)?
For example 48 ALU in 2x24 pipeline or 3x16 pipeline.
If the rumoured R500 Xenon GPU will have 48 unified ALUs from the leaked specs and is suspected to be manufactured on 90nm process, then I would expect the nVidia GPU to have 64 ALUs on a 65nm process and with a good dose of eDRAM. :) ...And possibly 2x32 pipeline configurations...though I couldn't tell you the benefits of 1x64, 2x32, 4x16 for what they would want to achieve with CELL, any guesses?
I would think these ALUs would be unified cores that would be more fragment/pixel oriented as the CELLs cores already have units suitable for vertex processing. :)
Could it be that next gen differentiator be that Cell architeture would allow ray tracing/Reyes and Xenon next system a more traditional approach to shadeing yet with much more performance than todays gpu, per pixel lighting, etc...?
AFAIK, there will be low level access granted to devs on both systems and both architectures should be flexible enough to try these things. The difference seems to be both the PS3 CPU (CELL) and the nVidia GPU will be stream processors but only the XeGPU will be a stream processor. What this means to developers, I couldn't tell you but they'll sure have fun finding out! ;)
From what we know of guestimates from Xenon leaked diagrams and recent CELL announcements,
Xe CPU @ 3.5 GHz ~ 84 Gflops
PS3 CPU @ 4.6 GHz ~ 294 Gflops
PS3 CPU 3.5x > Xe CPU
And if Xe GPU is on 90 nm process and PS3 GPU is on 65 nm process you would expect a doubling in transistor counts for a given die size on 65nm. So from Xenon leak,
Xe GPU = 240 GFlops
PS3 GPU 2x > Xe GPU
So as a very rough scaling calculation,
PS3 (CPU=294 + GPU=480) 2.4x > Xe (CPU=84 + GPU=240)
However, this is just a guestimate on peak performance and there are may other factors innvolved. Also Xe could have more cores or PS3 more CELLs/ PEs etc by the time of release. And no flames please, it's just a rough guestimate! ;)
If this is any indication then the extra power could translate into effectively trying new things as REYES/raytracing etc. but what I'm most interested in is how that would translate into gameplay and image quality on screen for games! :P The only thing I have to go on is the ~100 Gflops Sony GScube (+ SGI origin onyx) and if you we're impressed by that in 2001, then sure as heck, you'd be impressed by these consoles next gen for ~ $300! ;)
Raytracing on GPUs seem like a dead end to me.
The benefit of raytracing is in the way it allows a more correct handling of refraction and specular reflection. This makes it ill suited for a stream based approach because it introduce a data dependent step that has, potentially, highly irregular access patterns (cache busting).
I can see GPUs as accelerating first intersection rays (which is also what the above paper deals with), but these account for just a fraction of the work a heavy duty raytracer has to do.
Cheers
Gubbi
That paper mentions that by using tile rendering you can use the cache more efficiently. Perhaps it's time for nVidia to introduce Gigapixels TBDR technology into the GPU :?: 8)
It still only deals with first intersection rays.
The problem is exactly the same for CELL; poor random memory access.
Cheers
Gubbi
Megadrive1988
13-Dec-2004, 23:13
AFAIK, there will be low level access granted to devs on both systems and both architectures should be flexible enough to try these things. The difference seems to be both the PS3 CPU (CELL) and the nVidia GPU will be stream processors but only the XeGPU will be a stream processor. What this means to developers, I couldn't tell you but they'll sure have fun finding out! Wink
From what we know of guestimates from Xenon leaked diagrams and recent CELL announcements,
Xe CPU @ 3.5 GHz ~ 84 Gflops
PS3 CPU @ 4.6 GHz ~ 294 Gflops
PS3 CPU 3.5x > Xe CPU
And if Xe GPU is on 90 nm process and PS3 GPU is on 65 nm process you would expect a doubling in transistor counts for a given die size on 65nm. So,
PS3 GPU 2x > Xe GPU
So as a very rough scaling calculation,
PS3 (CPU + GPU) 7x > Xe (CPU + GPU)
However, this is just a guestimate on peak performance and there are may other factors innvolved. Also Xe could have more cores or PS3 more CELLs/ PEs etc by the time of release. And no flames please, it's just a rough guestimate! Wink
If this is any indication then the extra power could translate into effectively trying new things as REYES/raytracing etc. but what I'm most interested in is how that would translate into gameplay and image quality on screen for games! Razz The only thing I have to go on is the ~100 Gflops Sony GScube (+ SGI origin onyx) and if you we're impressed by that in 2001, then sure as heck, you'd be impressed by these consoles next gen for ~ $300! Wink
well, all we can say with near-certainty right now, is that Xe CPU will likely be sub 100 gflops or at most, around 100 Gflops. (maybe they've added a forth core or its always been a 4-core CPU as some reports indicated, i'll dig them up after work)
PS3 CPU is likely to be in the 100s of Gflops peak performance. whether it's the low 100s or upper 100s, we don't know. the whole 1 Tflop thing is so unimportant at this stage, i don't even want to think about that anymore. it's enough to say that PS3's successor in the early part of the next decade, will be a multi Tflops machine.
so it's looking likely that PS3 CPU will outclass Xe CPU by several times in peak performance. it will all come down to what the VPU/GPU in Xe and PS3 can output to the screen anyway, and there, things are likely to be more equal (ATI-MS vs Nvidia-Sony), that is, if the PS3 GPU is heavily based on nextgen GeForce architecture. I could well be wrong, and Vince suggests that those thinking that PS3 GPU will be like typical PC GPU architecture are dead wrong. it could well be that Sony is cooking up a 'GeForceSynthesizer' or 'Cell-Force' GPU, with ALL of the strong points of the GS family (GS, GS I-32, GS2, GS3/Visualizer) plus all of the strong points of Nvidia's nextgen GPU.
btw, the GSCube came out in 2000, iirc.
Thanx Jaws.
The only thing I have to go on is the ~100 Gflops Sony GScube (+ SGI origin onyx) and if you we're impressed by that in 2001, then sure as heck, you'd be impressed by these consoles next gen for ~ $300!
I remeber the GScube presentation in 2001 @ SIGGRAPH. However, given some of the comments Sony stated about the GScube (they wanted more performance) ain't it more likely that with Cell/Nvidia that performance targets are around GScube 64 (realistic rendering in realtime).[/quote]
Megadrive,
BTW, I've edited my above post slightly for one CELL, in GFlops,
PS3 (CPU=294 + GPU=480) 2.4x > Xe (CPU=84 + GPU=240)
Btw, it was both in 2000 and 2001, iirc. The latter had Square render FFSW also on nVidia hardware. Both Sony's and nVidia strategy/R&D has been to bridge the gap between offline rendering and realtime rendering. And this patent on Exluna rendering tech would be neat for a hardware implementation. i think both nVidia's and Sonys goals could be met with the PS3. Perhaps nVidia was restricted by MS/DX10/DXNext and Sony complememted it's future direction better?
Mythos,
If GScube 64 ~ 400 GFlops (not sure on SGI but they had 32 Mips R1400 cores, iirc), so 600-800 GFlop prolly in total (I'll add Mips if I find the figures...), then as in above, one CELL + GPU ~ 800 GFlops. So it would seem to fall in the same ballpark as a GScube-64 but it would have way less memory. :)
...
so it's looking likely that PS3 CPU will outclass Xe CPU by several times in peak performance. it will all come down to what the VPU/GPU in Xe and PS3 can output to the screen anyway, and there, things are likely to be more equal (ATI-MS vs Nvidia-Sony), that is, if the PS3 GPU is heavily based on nextgen GeForce architecture. I could well be wrong, and Vince suggests that those thinking that PS3 GPU will be like typical PC GPU architecture are dead wrong. it could well be that Sony is cooking up a 'GeForceSynthesizer' or 'Cell-Force' GPU, with ALL of the strong points of the GS family (GS, GS I-32, GS2, GS3/Visualizer) plus all of the strong points of Nvidia's nextgen GPU.
Performance between Xe GPU and PS3 GPU will largely depend on the Process used and die size (same would apply to CPUs, although CELL has an inherent advantage from a clean design). If equal in die size but Xe GPU is on 90nm and PS3 GPU is 65nm, roughly scaling, you should be able to get twice as many transistors at 65nm that would roughly equate to twice the performance.
My current thinking is that from Hofstee's presentation,
CPU<=>GPU--->output
This CPU|GPU logic can be on one IC and can scale to many ICs or you can split the CPU and GPU portions to separate ICs and scale that way (i.e PS3). We should see SLI, distributed rendering tech somewhere inherently built into the GPU to complement the ditributed processing of CELL. This would also lead me to beleive that the GPU would be CELL based and run software Cells.
I would think that the GPU would be some sort of unified ALUs but more focused for fragment/pixel processing as CELL is already a vertex powerhouse. And maybe you could load balance software Cells between CPU and GPU. In fact if you think of a current GPU with vertex and fragment processors but sclae it up, it should resemble the entire PS3 CPU(vertex)+GPU(fragment) chipset! :shock: However, these components in the GPU could be mixed Sony-Nvidia IPs, e.g Salc/Salps or pixel engines etc.
Also a good chance to see Gigapixel, TBDR to complement the parallel brick/tile rendering SCE patent! :P
We should see some Gelato rendering tech from this patent by Larry Gritz somewhere with REYES and raytracing/global illumination enabled shaders. 8)
And maybe finally they may roll out the Voodoo brand name...Voodoo Cells anyone! :P
http://online.cs.nps.navy.mil/DistanceEducation/online.siggraph.org/2002/Papers/13_GraphicsHardware/purcell.ppt
It works for me...it's a direct link...PM me if it still doesn't work...
I got it here if anyone wants
http://cggmwww.csie.nctu.edu.tw/seminar/GPU/2003.11.06.jhlu.ray.tracing.on.programmable.graphi cs.hardware.ppt
Interesting R8500 is better than GF3, I am right?
It says using limited fixed function though...the R8500 was able to use its fixed function abilities along with its dx8.1 abilities.
Megadrive1988
14-Dec-2004, 00:54
that's alot of stuff to think about Jaws. I don't really disagree, but then I don't really know either.
being slightly pessimistic (and cautious)
I believe that Sony will at least reach the goals it had for PS2, in the PS3. that is, realtime PS1 CG in games. maybe even somewhat beyond PS1 CG but I should not say one way or another since I don't know.
and then Sony should be able reach the goals it had for PS3, in the PS4. i.e. multi Tflop performance + raytracing + global illumination + full hardware accelerated software rendering, 1000 or 1000s of times more performance than PS2. etc.
Although I am now counting on a very very 'solid' PS3. maybe not with as much performance as we once thought or wanted (yes PS3 will be very powerful) yet, with much much greater visual quality (Nvidia) and relative ease of development (better than even PSP) for developers who just want to get started. PS3 will be better balanced overall than I once thought (before Nvidia was there).
yes it will be difficult to make full use of the PS3 CPU for some years, but it will get there. the situation will be much better than the early days of PS2, and even now with PS2. although this says *nothing* about the amount of resources and manpower needed to make stunning large scale nextgen games. that will still be emense.
DaveB, Vince and Pana, do you guys disagree with what I'm saying?
A little off the subject...
But IF Cell/GPU reach Upper level Flops with Ray tracing, GI, etc..We may see a new cross pollunation with a version of a game containing a short film (or other forms of interactivity). War devil has mention this, movement of EA to LA studios.
and then Sony should be able reach the goals it had for PS3, in the PS4. i.e. multi Tflop performance + raytracing + global illumination + full hardware accelerated software rendering, 1000 or 1000s of times more performance than PS2. etc.
Before PS4, let's make an SLI-ed PS3 8) A PS3 should be with an optical interface, so...
Megadrive1988
14-Dec-2004, 02:42
sure, i would love SLI PS3s. stack a few, or many PS3s together. combined their processing power (not over the internet).
Performance between Xe GPU and PS3 GPU will largely depend on the Process used and die size (same would apply to CPUs, although CELL has an inherent advantage from a clean design). If equal in die size but Xe GPU is on 90nm and PS3 GPU is 65nm, roughly scaling, you should be able to get twice as many transistors at 65nm that would roughly equate to twice the performance.
First, how do you know that Xenon CPU isn't a clean design?
Secondly, how do you know there is a process advantage?
CPU<=>GPU--->output
This applies equally to Xenon as much as PS3, Xenon has been built with good connections between CPU and GPU.
This CPU|GPU logic can be on one IC and can scale to many ICs or you can split the CPU and GPU portions to separate ICs and scale that way (i.e PS3). We should see SLI, distributed rendering tech somewhere inherently built into the GPU to complement the ditributed processing of CELL. This would also lead me to beleive that the GPU would be CELL based and run software Cells.
I would think that the GPU would be some sort of unified ALUs but more focused for fragment/pixel processing as CELL is already a vertex powerhouse. And maybe you could load balance software Cells between CPU and GPU. In fact if you think of a current GPU with vertex and fragment processors but sclae it up, it should resemble the entire PS3 CPU(vertex)+GPU(fragment) chipset! :shock: However, these components in the GPU could be mixed Sony-Nvidia IPs, e.g Salc/Salps or pixel engines etc.
Also a good chance to see Gigapixel, TBDR to complement the parallel brick/tile rendering SCE patent! :P
We should see some Gelato rendering tech from this patent by Larry Gritz somewhere with REYES and raytracing/global illumination enabled shaders. 8)
And maybe finally they may roll out the Voodoo brand name...Voodoo Cells anyone! :P
Is it usefull to have a "next generation" Hardware with a geometric stage separated to rasterization phase? Vertex in the Cell CPU and fragment/pixel in the NVidia GPU?
What kind of connection between Cell and GPU, and i think it would not be best to have a unified 'shader core' in a GPU to perform vertex+pixel operatrion?
version
14-Dec-2004, 10:46
This CPU|GPU logic can be on one IC and can scale to many ICs or you can split the CPU and GPU portions to separate ICs and scale that way (i.e PS3). We should see SLI, distributed rendering tech somewhere inherently built into the GPU to complement the ditributed processing of CELL. This would also lead me to beleive that the GPU would be CELL based and run software Cells.
I would think that the GPU would be some sort of unified ALUs but more focused for fragment/pixel processing as CELL is already a vertex powerhouse. And maybe you could load balance software Cells between CPU and GPU. In fact if you think of a current GPU with vertex and fragment processors but sclae it up, it should resemble the entire PS3 CPU(vertex)+GPU(fragment) chipset! :shock: However, these components in the GPU could be mixed Sony-Nvidia IPs, e.g Salc/Salps or pixel engines etc.
Also a good chance to see Gigapixel, TBDR to complement the parallel brick/tile rendering SCE patent! :P
We should see some Gelato rendering tech from this patent by Larry Gritz somewhere with REYES and raytracing/global illumination enabled shaders. 8)
And maybe finally they may roll out the Voodoo brand name...Voodoo Cells anyone! :P
Is it usefull to have a "next generation" Hardware with a geometric stage separated to rasterization phase? Vertex in the Cell CPU and fragment/pixel in the NVidia GPU?
What kind of connection between Cell and GPU, and i think it would not be best to have a unified 'shader core' in a GPU to perform vertex+pixel operatrion?
between CPU and GPU is REDWOOD connection it is fast 25-100 GB/s,but texture upload too slow if polygon datas send CPU for GPU
nvidia say ps3 performance 50 *ps2= 3 gigapolygon/s
1 vertex about 16 byte(xyz,color,uv), 3 Ggpoly/s=50 GB/s
if REDWOOD 100 GB/s then texture upload is fine , otherwisw is shit
london-boy
14-Dec-2004, 11:06
Now, this might be an itsy pitsy little OT, but i have a question for you people.
If the PS3 GPU is based on NVIDIA's next gen architecture (NV50), why would NVIDIA cancel the NV50 for PC? Doesn't make sense.
version
14-Dec-2004, 14:41
Now, this might be an itsy pitsy little OT, but i have a question for you people.
If the PS3 GPU is based on NVIDIA's next gen architecture (NV50), why would NVIDIA cancel the NV50 for PC? Doesn't make sense.
1. new type of pixelpipeline?
2. embended memory is very fast , probably it ?
3. raytracing? not believe 10 teraflops the minimum for it
3. go out pc biznis:) ?
london-boy
14-Dec-2004, 14:44
The thing is, i don't see why they would cancel their PC product if their PS3 GPU will be based on the same architecture!!
Of course it's just a rumour, but even as a rumour it sounds strange.
It's not like they haven't got the capabilities to work on more than one chip at the same time... Especially if the 2 chips are similar...?
version
14-Dec-2004, 14:53
The thing is, i don't see why they would cancel their PC product if their PS3 GPU will be based on the same architecture!!
Of course it's just a rumour, but even as a rumour it sounds strange.
It's not like they haven't got the capabilities to work on more than one chip at the same time... Especially if the 2 chips are similar...?
pc will run ps3's games ?, NO, sony not idiot
london-boy
14-Dec-2004, 15:04
The thing is, i don't see why they would cancel their PC product if their PS3 GPU will be based on the same architecture!!
Of course it's just a rumour, but even as a rumour it sounds strange.
It's not like they haven't got the capabilities to work on more than one chip at the same time... Especially if the 2 chips are similar...?
pc will run ps3's games ?, NO, sony not idiot
What are you on about?
It's not like Sony are gonna force NVIDIA to stop producing parts for PC for fear that PCs might be able to run PS3 games! :lol: :roll:
Inane_Dork
14-Dec-2004, 19:24
If the PS3 GPU is based on NVIDIA's next gen architecture (NV50), why would NVIDIA cancel the NV50 for PC? Doesn't make sense.My guess: PS3 and pleasing Sony is about 1000000x more important than coming out with another $500 PC card.
As for why they dropped the NV50, it likely has to do with how different their PS3 chip will be. Cut down or removed vertex processing. More cache/embedded RAM. Only chip needed (not entire card).
I don't think there's a single nvidia investor that would agree with your statement (well there's may be one or two, but they are probably crazy.). PS3 is important, but not more important then their core PC oriented business. I could swear I read some analyst mention that this deal could potentially be less money compared to the xbox deal (since it's a tech licensing deal and they aren't selling chips directly).
If I recall, when Nvidia provided a graphics chip for Xbox, they too delayed the launch of new technology in favour of a geforce 3 refresh. Then they rolled the tech from the XGPU back into the geforce 4 (with what little differences exsisted) the following year.
The thing is, i don't see why they would cancel their PC product if their PS3 GPU will be based on the same architecture!!
Of course it's just a rumour, but even as a rumour it sounds strange.
It's not like they haven't got the capabilities to work on more than one chip at the same time... Especially if the 2 chips are similar...?
IMO, I think that NVidia has changed directions on where they wanted to be with their next PC GPU. They must've realized that they where going down the same path (more performance/some additoinal features) with NV48/NV50. At the same time they where working with Sony and closed the deal on the GPU work thus allowing them to use some SONY IP (maybe Toshiba). So Nvidia realized they can get better performanceing and in addition bridging this generation rendering with next gen rendering techniques of possibly REYES, Ray tracing. Possibly PS3 will represent a main focus techinque of REYES, ray traceing, GI..etc.[/quote]
You really think ray tracing is gogint o be a big part of the PS3's graphics capabilites?
I dunno, I just think it's a step beyond what anyone should expect in a console for next generation.
Megadrive1988
14-Dec-2004, 20:20
does anyone here believe that raytracing, global illumination, and those other almost-impossible-to-do-now-in-realtime (on current conventional hw) types of graphics will not really be done on PS3, but more like PS4 ?
edit: seems that Qroach and I have a similar question :)
I think the possibilty of some form of ray tracing is more than likely...Examples: Sony's pantent in a form of ray tracing is used, unreal engine 3 is to have some form of ray tracing.
hey we did have the same question :)
Well unreal engine 3 is design around PC style hardware (obviously since where else are you going to develop with pixel shaders ). However where did you read it uses raytracing?
NO WAY they're going to do ray-tracing. Not only hardware and memory access is still too slow to do RT with the number of passes needed to produce great looking images, but it would also mean radical changes in gpu/development tools. I just can't see nVIDIA making such a radical change in their new line of GPU's.
Well, technically that is the direction Nvidia is heading. Whether it will happen any time soon is the real question.
GwymWeepa
14-Dec-2004, 20:47
I don't think there's a single nvidia investor that would agree with your statement (well there's may be one or two, but they are probably crazy.). PS3 is important, but not more important then their core PC oriented business. I could swear I read some analyst mention that this deal could potentially be less money compared to the xbox deal (since it's a tech licensing deal and they aren't selling chips directly).
If I recall, when Nvidia provided a graphics chip for Xbox, they too delayed the launch of new technology in favour of a geforce 3 refresh. Then they rolled the tech from the XGPU back into the geforce 4 (with what little differences exsisted) the following year.
A single card doesn't even come close to the revenue a single console contract can bring in, Xbox revenues amounted to something like 20% of Nvidias revenue for a long time...so having a long-term contract I think is more important than any single card, they can always rebound, its not like they are quitting PC development for 5 years as the ps3 has its run.
I have no doubt that RT is the future. I bet the guys at ATI/nVidia are dying to produce something fast enough to do a realtime rt. But IMO, it will only be worth while when they be able to do real good quality raytracing. Otherwise, shaders will still beat RT by far in image quality. Right now, most realtime raytraced images are pretty plain and ugly.
A single card...
That sells for how many years and at what price? I wouldn't be suprised if Nvidia made more money off a single card, then they did from all the components supplied for a single xbox.
doesn't even come close to the revenue a single console contract can bring in, Xbox revenues amounted to something like 20% of Nvidias revenue for a long time...
Like I said, This deal with sony is a licensing deal, not the same as MS buying fabbed chips from Nvidia, where Nvidia can control the price and factor in profits. Nvidia will obviously get a royaltee but I don't think that will be more than they would get from a single video card sale, or the high price MS "had" to pay.
I have no doubt that RT is the future. I bet the guys at ATI/nVidia are dying to produce something fast enough to do a realtime rt. But IMO, it will only be worth while when they be able to do real good quality raytracing. Otherwise, shaders will still beat RT by far in image quality. Right now, most realtime raytraced images are pretty plain and ugly.
A quick post just before I pop out, read the title of this thread...and the patent...It's a shader implementation for REYES that also enables raytracing/GI algorithms to be applied and optimised! ;)
However where did you read it uses raytracing?
A form of ray tracing..."Normal-mapped translucent object distorts and attenuates the frame buffer, simulating ray-traced reflections."
unrealtechnology.com
Also, check this out.... "In summary, we were able to support all the traditional effects of Quake 3 while mosteffects were significantly simpler to implement. Looking at newer engines such as Unreal3, we still see no effects that could not be supported easily by ray tracing."
http://216.239.63.104/search?q=cache:YltN3g1ECNEJ:graphics.cs.uni-sb.de/~jofis/Arbeit/GI04-MWZC-RayTracing.pdf+unreal+engine+3+ray+tracing&hl=en
[/quote]
Fafalada
14-Dec-2004, 21:10
A form of ray tracing..."Normal-mapped translucent object distorts and attenuates the frame buffer, simulating ray-traced reflections."
This "form of rayracing" has been used in games since at least ~1995, if not earlier.
Qroach:
Which graphics card has been sold 100 million times so far?
Which graphics card has been sold 100 million times so far?
How often do people or system vendors purchase graphics cards? How many different types of graphics cards are on the market at any one time?
Megadrive1988
14-Dec-2004, 21:33
I'm pretty confident that we'll see RT raytracing, radiosity, GI, etc in future consoles. X3-N6-PS4 if not X4-N7-PS5. I think that X2-N5-PS3 are going to be far too slow and weak (ha!) to acomplish these advanced rendering techniques in RT in games, and have complex visuals on top of it.
Which graphics card has been sold 100 million times so far?
How often do people or system vendors purchase graphics cards? How many different types of graphics cards are on the market at any one time?
I'm sure you have the numbers... enlighten me. :wink:
on the other hand, did you also factor in that Sony will be ultimately be fabbing these chips once Nvidia's work is done? Did you factor in how much work Nvidia will be doing in the lifespan the PS3 is expected to sell and how much money on royalties will be made in that time... :oops:
on the other hand, did you also factor in that Sony will be ultimately be fabbing these chips once Nvidia's work is done? Did you factor in how much work Nvidia will be doing in the lifespan the PS3 is expected to sell and how much money on royalties will be made in that time...
Yes I did. If Nvidia doesn't fab the chips and sell them, they won't make a have to spend as much money, but they also won't make nearly as much profit on each chip. The bottom line I think some anylysts out there saw was that even though sony is a good console to be invovled with, it wouldn't necessarily be worth more money compared to previous deals OR their core PC business in the long run.
I think how much this deal is worth has already factored into the price of Nvda stock.
GwymWeepa
14-Dec-2004, 22:38
A single card...
That sells for how many years and at what price? I wouldn't be suprised if Nvidia made more money off a single card, then they did from all the components supplied for a single xbox..
Oh Nvidia probably makes much more money off a single card sold than a single xbox, and they'll make a lot more off a high-end chip offered now than a single ps3...but when those consoles sell into the multiple millions, that's when the real money is made. It wouldn't surprise me if the ps3 sells over 100 million units, that's a lot of f'ing cash to be made, far more than any single card. A hiccup in their pc line-up when designing the ps3 chip is nothing, especially when factoring in how much money they'll make from the console contract.
The bottom line I think some anylysts out there saw was that even though sony is a good console to be invovled with, it wouldn't necessarily be worth more money compared to previous deals OR their core PC business in the long run.
It's about corporate image, rather than immediate money. Do you trust some obscure analyst, or the assessment done by Sony towards the value of nVIDIA technology? Also nVIDIA's contribution is not limited in hardware IP - it includes software and support which will continue at least several years. In addition the relationship with Sony has importance in OpenGL community, as those with actual products have big voice in there anyway. nVIDIA is joining in the mobile market too, where OpenGL ES is standardized, not to mention the consumer electronics area. It has a potential of the nVIDIA breakout into whole new markets. Compared to this, the PC card market is just diminishing now - the battle between cheap videocard-integrated chipsets.
I'll trust people looking at what this is really worth in dollars and cents as opposed to PR released by Sony or Nvidia.
Nvidia always supports hardware they create with software and tools. no suprise here... They do it on the PC, did it for xbox, and now ps3. Nvidia was targetting the expansion into new markets long before any deal with sony was struck. Remember when they said (years ago) about having a graphics processor driving every screen?
I think Nvidia moving into new markets was an evoloutionary step, not revoloutionary. Nvidia already has the biggest name in graphics next to ATI, just how is thier image going to improve beyond that? To be honest, I think you're really stretching to find reason that support your opinion.
on the other hand, did you also factor in that Sony will be ultimately be fabbing these chips once Nvidia's work is done? Did you factor in how much work Nvidia will be doing in the lifespan the PS3 is expected to sell and how much money on royalties will be made in that time...
Yes I did. If Nvidia doesn't fab the chips and sell them, they won't make a have to spend as much money, but they also won't make nearly as much profit on each chip. The bottom line I think some anylysts out there saw was that even though sony is a good console to be invovled with, it wouldn't necessarily be worth more money compared to previous deals OR their core PC business in the long run.
I think how much this deal is worth has already factored into the price of Nvda stock.
Nvidia being responsible for the fabbing costs money too. In turn, they can of course sell them at a even higher price (which is ultimately what you're refering to and what they did with the Xbox deal) - yet, simply designing the chip has its distinct advantages in that it's a one-off effort and will ultimately be paying off over a potential lifespan of 7 years. Think about it again: designing a chip with a one-off effort together with Sony and you'll be enjoying royalities for a console with a potential 100+ million userbase after 5 years. Not only that, but as One mentioned, market presence in also a factor that will pay off, if only by mindshare.
Working together with the number 1 console maker is far more lucrative than you think - even if it's just a licencing contract. One shouldn't forget that while Nvidia had the upperhand with Xbox and the asking price for their chips - the Xbox is rather far behind in userbase numbers and that doesn't even include the negative mindshare of having your parts in the "financially disasterous xbox project".
Again, "not getting more than from a single video card sale" doesn't seem anywhere near plausible once you consider the factors involved and the little work/effort they're doing now in designing a chip that will last for years to come. Not to mention the experience and knowlege in being closely tied in with CELL related stuff that can and will be used in their future PC-orientated projects (at least it will be inspired to say the least).
Last factor to consider is the PC gaming sector that is on a decline and may/will continue to do so as console gaming becomes even more popular and more mainstream. It simply is a good thing for them to be on this deal and making an appearance... much like Microsoft is doing with Xbox as well.
I'll trust people looking at what this is really worth in dollars and cents as opposed to PR released by Sony or Nvidia.
PR? LOL, The fact 'Sony chose nVIDIA' is THE approval. Sony engineers can assess technical merits in nVIDIA technology better than financial analysts in Wall streets, don't you agree?
Nvidia always supports hardware they create with software and tools. They do it on the PC, did it for xbox, and now ps3. Nvidia was targetting the expansion into new markets long before any deal with sony was struck. Remember what they said years ago about having a graphics processor driving every screen?
I think Nvidia moving into new markets was an evoloutionary step, not revoloutionary. Nvidia already has the biggest name in graphics next to ATI, just how is thier image going to improve beyond that? To be honest, I think you're really stretching to find reason that support your opinion.
Eh? Can you name other big clients of nVIDIA, except for Sony and Microsoft? Evolutionary or revolutionary, it's a big plus for nVIDIA. nVIDIA is a big name in the PC space, but frankly its brand name is obscure in other markets at this time.
PR? LOL, The fact 'Sony chose nVIDIA' is THE approval. Sony engineers can assess technical merits in nVIDIA technology better than financial analysts in Wall streets, don't you agree?
I haven't been talking about "technical" I've been talking about "financial".
Eh? Can you name other big clients of nVIDIA, except for Sony and Microsoft? Evolutionary or revolutionary, it's a big plus for nVIDIA. nVIDIA is a big name in the PC space, but frankly its brand name is obscure in other markets at this time.
Nvidia is a partner just like IBM and toshiba in a sense. If you really think sony isn't going to plaster Nvidia's name on the PS3 or even refer to them by name nearly as much as microsft did, then I think you'll be shocked. I won't be suprised if you don't even see Nvidia's name in anywhere but the hardware specs (if even that).
PR? LOL, The fact 'Sony chose nVIDIA' is THE approval. Sony engineers can assess technical merits in nVIDIA technology better than financial analysts in Wall streets, don't you agree?
I haven't been talking about "technical" I've been talking about "financial".
There must be hidden assets, which are future-proof, in nVIDIA. I bet Sony people under NDA could evaluate them and reached the conclusion. If Sony people are all stupid, then I'm wrong.
Eh? Can you name other big clients of nVIDIA, except for Sony and Microsoft? Evolutionary or revolutionary, it's a big plus for nVIDIA. nVIDIA is a big name in the PC space, but frankly its brand name is obscure in other markets at this time.
Nvidia is a partner just like IBM and toshiba in a sense. If you really think sony isn't going to plaster Nvidia's name on the PS3 or even refer to them by name nearly as much as microsft did, then I think you'll be shocked. I won't be suprised if you don't even see Nvidia's name in anywhere but the hardware specs (if even that).
Why don't you think partnership is not enough? I'm not talking about building consumer brands. When some cellphone maker evaluates imaging hardware, it evaluates nVIDIA technology by cost and technical merits, then cuts a deal. In evaluating cost and technical merits, the PS3 deal can be one of factors. Consumer perception is irrelevant.
Performance between Xe GPU and PS3 GPU will largely depend on the Process used and die size (same would apply to CPUs, although CELL has an inherent advantage from a clean design). If equal in die size but Xe GPU is on 90nm and PS3 GPU is 65nm, roughly scaling, you should be able to get twice as many transistors at 65nm that would roughly equate to twice the performance.
First, how do you know that Xenon CPU isn't a clean design?
By 'clean' if you mean efficient, custom CPU design, then yes, I have no doubt, it will be a great multi core CPU compared to curent CPUs with neat custom features. Heck, I also have a Mac and if someone told me a triple core G5 @ 3.5Ghz ~ $300 would be available Q4 2005 in a console, I would be impressed.
However by 'clean', I was referring to a 'clean sheet' design, starting with a blank sheet of paper and 400 engineers at your disposal from STI and dedicated design centre type of 'clean'. If they haven't f%£ked, it should inherently perform better for a given process and die size. This was always STI's goal stated throughout the project.
Secondly, how do you know there is a process advantage?
Well, with the scaling calcs I did earlier with leaked specs etc,
Xe CPU(tri-core) = 84 GFlops @ 3.5GHz, 65nm (IBM)
Xe GPU = 240 GFlops @ 500Mhz, 90nm (TSMC)
PS3 CPU(1 CELL) = 294 GFlops @ 4.6GHz, 65nm (STI)
PS3 GPU = 480 GFlops @ 500MHz, 65nm (STI)
Of course, change clocks, cores, process etc. accordingly scale depending on your POV. These process assumptions were based on existing discussions on these boards and what was expected in the release timeframe. However reading this below,
http://eetimes.com/semi/news/showArticle.jhtml?articleID=52601132
There's a chance that TSMC will have 65nm ready for late 2005. If that's the case then the GPUs should be similar in performance for a given die size. However, I would be extremely impressed if Xe is totally at 65nm for Q4 launch and matches PS3 at 65nm for H2 2006 launch and would suggest questions at STIs fab strategies and heavy investments to be ahead of the curve. But that would also hint at PS3 launching on 45nm in 2006 as Sony+Tosh have stated to have 45nm sampling by Q4 2005.
http://www.sony.net/SonyInfo/News/Press/200402/04-0212E/
CPU<=>GPU--->output
This applies equally to Xenon as much as PS3, Xenon has been built with good connections between CPU and GPU.
Yes, this was inferred in the leaked specs with L2 cache being read by GPU etc. but I didn't mean to imply the contrary. If you followed my post, I was trying to show two different ways to scale CPU+GPU, i.e. where [...]=IC,
Onchip IC,
[CPU<=>GPU]--->Output
Separate chip IC's (PS3),
[CPU]<=>[GPU]--->Output
Just curious for opinions, which RAM config would be better for PS3,
A) Off CPU,
[CPU]<=>[GPU]--->Output
|
[RAM]
B) Split CPU|GPU,
[CPU]<=>[GPU]--->Output
| |
[RAM] [RAM]
Say for a total of 512MB which would be better, A or B :?:
Nvidia being responsible for the fabbing costs money too. In turn, they can of course sell them at a even higher price (which is ultimately what you're refering to and what they did with the Xbox deal) - yet, simply designing the chip has its distinct advantages in that it's a one-off effort and will ultimately be paying off over a potential lifespan of 7 years. Think about it again: designing a chip with a one-off effort together with Sony and you'll be enjoying royalities for a console with a potential 100+ million userbase after 5 years. Not only that, but as One mentioned, market presence in also a factor that will pay off, if only by mindshare.
Working together with the number 1 console maker is far more lucrative than you think - even if it's just a licencing contract.
Really, then why hasn't the stock price for nvidia shot through the roof? I'm only talking about lucrative in a financial sense. I have no interest in arguing about "perception".
One shouldn't forget that while Nvidia had the upperhand with Xbox and the asking price for their chips - the Xbox is rather far behind in userbase numbers and that doesn't even include the negative mindshare of having your parts in the "financially disasterous xbox project".
Regardless of how many xbox consoles were sold, Nvidia stiull made a $hit load of money off it, and still does. Negative mindshare? Um, in a financial sense there is no such thing. Mindshare doesn't make you money or else 3dfx would still be the #1 graphics chip maker.
Again, "not getting more than from a single video card sale" doesn't seem anywhere near plausible once you consider the factors involved and the little work/effort they're doing now in designing a chip that will last for years to come.
Little work? so which is it that you want to argue for? on one side you argue they are providing tools and software and on the other you argue little work/effort. The two really don't go together. providing tools and software is a lot fo continuous work, once you can't describe as little "work/effort". Nvidia always designs a chip that will last for years to come.
Last factor to consider is the PC gaming sector that is on a decline and may/will continue to do so as console gaming becomes even more popular and more mainstream. It simply is a good thing for them to be on this deal and making an appearance... much like Microsoft is doing with Xbox as well.
Video cards or GPU's are only required for video games? No of course not. If that was the case expanding into other markets such as you mentioned wouldn't make sense right? This goes back to my previous point, regarding the average life of graphics processors and the sheer amount available on the market at one time. Nvidia made tons of money of TNT processors despite them being incredbly out of date.
Mindshare doesn't make you money or else 3dfx would still be the #1 graphics chip maker.
I bet what 3dfx got was mere 'geek mindshare' :lol:
You have to note that the PC vidcard market is very fast in cycling products.
It's very risky.
There must be hidden assets, which are future-proof, in nVIDIA. I bet Sony people under NDA could evaluate them and reached the conclusion. If Sony people are all stupid, then I'm wrong.
I don't think there's anythign such as "future proof" in technology these days. anyway I'm sure Sony is getting a really good deal and good technology form nvidia. Why is it that whenever soeone has an alternate explanation for somehting invovling sony and PS3, they always resort to, "well sony isn't stupid" or "if you think sony is stupid, then..." when
I've said nothing of the sort. I've been talking about what the sony deal is worth in financial term to Nvidias bottom line. I haven't been saying anything remotely like sony is stupid for using nvidia or anything like that.
When some cellphone maker evaluates imaging hardware, it evaluates nVIDIA technology by cost and technical merits, then cuts a deal. In evaluating cost and technical merits, the PS3 deal can be one of factors. Consumer perception is irrelevant.
That's actually what I've been trying to convey. consumer perception is irrelevant in this deal.
yes PC video cards is risky, which is why Nvidia and ATI release cards targetting all the market segments out there.
http://eetimes.com/semi/news/showArticle.jhtml?articleID=52601132
There's a chance that TSMC will have 65nm ready for late 2005.
According to the article
This year, TSMC will spend about US$400 million on process development, with most of that going to development of the 65-nm node. Earlier this year, the company made its first 65-nm SRAM test module and announced plans to make the first low-power devices at that process node by late 2005. A high speed version of the process is scheduled to come in the first half of 2006, followed by a general purpose process later that year.
If the Xenon release is in late 2005, it's unlikely 65nm lines will be ready then.
As for XeCPU, reading articles like this,
http://www.theregister.co.uk/2004/02/02/first_65nm_ibm_powerpc_chip/
I doubt it is in 65nm.
BTW, the latest devkits for Xbox 2 are still in a dual-CPU Mac G5, right? Why don't they offer a 4-way version, like using 2 Xserve G5?
Mindshare doesn't make you money or else 3dfx would still be the #1 graphics chip maker.
I bet what 3dfx got was mere 'geek mindshare' :lol:
You have to note that the PC vidcard market is very fast in cycling products.
It's very risky.
Voodoo 3 was number 1 in the retail markets(probably due to its $100-$200 price range compared to nvidia's $200-$300, plus riding off the success of previous voodoos).
Voodoo 5, from its release until 3dfx shriveled up, outsold the combined sales in retail of the top 2 geforce 2 makers, which I assume means it was a close second in to Geforce 2 sales during that time.
randycat99
15-Dec-2004, 06:42
Really, then why hasn't the stock price for nvidia shot through the roof? I'm only talking about lucrative in a financial sense. I have no interest in arguing about "perception".
While I don't know what you consider "shooting through the roof", it does seem to be up 60% since about August, and is following a particularly healthy rise consistent with the recovery trend of the tech sector. It looks very good for them assuming they continue on this trend. While they may not be at levels comparable to past quarters (yet), you cannot ignore the troubles the entire tech sector has weathered since 2001- so everybody is starting from a low mark around now. The point is, they are rising, and rising considerably better than most tech stocks right now.
aaronspink
15-Dec-2004, 07:27
I think the possibilty of some form of ray tracing is more than likely...Examples: Sony's pantent in a form of ray tracing is used, unreal engine 3 is to have some form of ray tracing.
Unreal 3 won't and neither will PS3.
Aaron Spink
speaking for myself inc.
However by 'clean', I was referring to a 'clean sheet' design, starting with a blank sheet of paper and 400 engineers at your disposal from STI and dedicated design centre type of 'clean'. If they haven't f%£ked, it should inherently perform better for a given process and die size. This was always STI's goal stated throughout the project.
Yes I'm also refererring to a 'clean sheet' design.
Why do some people have a problem that XeCPU is a clean design, designed and built from scrarch for this one job. Blank sheet of paper, lots of engineers. As I have said many many times its not remotely related to a G5, its share a similar ISA but thats it and by the sounds of it so does Cell.
MS and IBM picked a set of features and built a processor around them.
Say for a total of 512MB which would be better, A or B :?:
From a developer point of view a unified memory system (excluding eDRAM and caches) is an advantage. Its also cheaper. Expect 256 Mb accessable from both CPU and GPU.
Li Mu Bai
15-Dec-2004, 10:19
Why is it that every patent filed by Toshiba/Sony/IBM somehow relates directly to the PS3? Especially when Cell has much further reaching technical implications for all parties involved? Reyes has been Pana's love child since any Cell patent was 1st filed, or the technology discussed. We all know that this amalgamation of companies didn't join together to solely create the PS3. I realize there's some merit in well-grounded intelligent speculation, but when the PS3 spec. sheet is finally revealed I can already foresee disappointment due to this type of thing. It's already been proven that the PS3 is not the behemoth that it was initially touted to be, but a beast nonetheless. Also, how many tech. patents are existing out there that have yet to come to fruition? (engineers will face headaches even in initial Reyes implementation, not to mention the external memory requirements for essentially a stream processor, ala the PS3 & its subpar random memory access for this type of full application.) Reyes has some time yet before it arrives & creates a paradigm shift in console or PC GPUs. (for many of the developers are one in the same, & this generation will see even more PC dev. cross-platform work, due to the more powerful console architectures & advanced feature sets identical to those of the high-end cards. Surpassing them even for a time) PS4, XBX2, & the NES6 is far more feasible for full Reyes utilization.
It will indeed be some time before the x86 domination of PCs comes to an end as well, unless you're underestimating the time frame in which MS wants/dictates the direction to advance or change. Unfortunately, they also possess the power & consumer mindshare currently to marginalize almost any OS application. Even if it is more efficient, etc. Gasp, I'm tired of reading these Reyes-centric threads. I can still remember when all of those early adherents of Sony's 1 teraflop power claims were defending its technical validity in a multitude of threads. :roll: Let's wait for more tangible evidence that this indeed is a real possibility, & not simply the first intersection rays if that.
london-boy
15-Dec-2004, 10:28
Why is it that every patent filed by Toshiba/Sony/IBM somehow relates directly to the PS3? Especially when Cell has much further reaching technical implications for all parties involved?
I think the real question is, why is it that it is so easy and understandable to "find" patents relating specifically to PS3, and not for other machines?
Maybe because people are kinda looking for patents relating to PS3?
not to mention the external memory requirements for essentially a stream processor, ala the PS3 & its subpar random memory access for this type of full application.)
IIRC XDR-DRAM is not bad in random access
http://pc.watch.impress.co.jp/docs/2004/0716/kaigai04.jpg
version
15-Dec-2004, 10:45
not to mention the external memory requirements for essentially a stream processor, ala the PS3 & its subpar random memory access for this type of full application.)
IIRC XDR-DRAM is not bad in random access
yeah, but xdr access 32 bit(32wire) , gddr3 have 256wire =256 bit about same time
gddr3 faster 8 times than xdr
yeah, but xdr access 32 bit(32wire) , gddr3 have 256wire =256 bit about same time
gddr3 faster 8 times than xdr
No way. XDR supports up to 128bit interface width (16 * 8-channel). Also, XDR is 2.5x faster per-pin than GDDR3 in 2005.
http://www.rambus.com/news/newsroom/pressrelease.cfm?id=110
XDR memory's novel system topology allows point-to-point differential data interconnects to scale to multi-gigahertz speeds, while the bussed address and command signals allow a scalable range of memory system capacity supporting from one to 36 DRAM devices. XDR offers a roadmap to 6.4GHz and can scale to interface widths of up to 128-bits, enabling memory system bandwidths up to 100GB/s, 16 times more than today?s 6.4GB/s memory systems. XDR DRAMs will be available in multiple speed bins, device densities, and device widths. With densities ranging from 256Mb to 8Gb, and device widths ranging from x1 to x32, XDR DRAM satisfies the needs of both high-bandwidth and high-capacity systems.
Samsung, Elpida, and Toshiba are all DRAM licensees of Rambus's latest memory interface technology.
version
15-Dec-2004, 14:17
yeah, but xdr access 32 bit(32wire) , gddr3 have 256wire =256 bit about same time
gddr3 faster 8 times than xdr
No way. XDR supports up to 128bit interface width (16 * 8-channel). Also, XDR is 2.5x faster per-pin than GDDR3 in 2005.
http://www.rambus.com/news/newsroom/pressrelease.cfm?id=110
XDR memory's novel system topology allows point-to-point differential data interconnects to scale to multi-gigahertz speeds, while the bussed address and command signals allow a scalable range of memory system capacity supporting from one to 36 DRAM devices. XDR offers a roadmap to 6.4GHz and can scale to interface widths of up to 128-bits, enabling memory system bandwidths up to 100GB/s, 16 times more than today?s 6.4GB/s memory systems. XDR DRAMs will be available in multiple speed bins, device densities, and device widths. With densities ranging from 256Mb to 8Gb, and device widths ranging from x1 to x32, XDR DRAM satisfies the needs of both high-bandwidth and high-capacity systems.
Samsung, Elpida, and Toshiba are all DRAM licensees of Rambus's latest memory interface technology.
:)
yes, xdr faster in burst mode, but with random access slow
ps3 have 32 or 64 bit(wire) for xdr
:)
yes, xdr faster in burst mode, but with random access slow
ps3 have 32 or 64 bit(wire) for xdr
No, the expected bandwidth for the PS3 memory configuration is
256MB XDR DRAM (3.2Gtps) (256Mbit * 8 @ 128bit) = 51.2GB/s
or
128MB @ 64bit for Cell + 128MB @ 64bit for GPU = 25.6GB/s + 25.6GB/s
while Xenon is
512MB GDDR3 (1.6Gtps) (512Mbit * 8 @ 256bit) = 51.2GB/s
or
256MB GDDR3 (256Mbit * 8 @ 256bit) = 25.6GB/s
Anyway, read the picture above "Optimized XDR core timings yield higher effective bandwidth under random workloads"
Little work? so which is it that you want to argue for? on one side you argue they are providing tools and software and on the other you argue little work/effort. The two really don't go together. providing tools and software is a lot fo continuous work, once you can't describe as little "work/effort". Nvidia always designs a chip that will last for years to come.
Little work compared to their on-going cyclic business in their core-market. Little work in a sense that they'll be living off revenues of a system that's is bound to sell 100+ million? For at least 6 years after completion and not having to do a thing, while they can happily concentrate on their core-market again? Little work in that they won't even be responsible for the chips fabbing but just will be designing the thing and then forgetting about it.
As for "Nvidia always designs a chip that will last for years to come" - I'm still interested to see actual numbers of a chip that sold in same league the PlayStation brand does and is excected to continue in its third iteration. If there isn't - it just doesn't cut it. In fact, if their core-business was that lucrative, they wouldn't be doing a licencing deal with Sony, obviously. Having said that, Nvidia's actions speak far more than your arguments do...
version
15-Dec-2004, 15:43
:)
yes, xdr faster in burst mode, but with random access slow
ps3 have 32 or 64 bit(wire) for xdr
No, the expected bandwidth for the PS3 memory configuration is
256MB XDR DRAM (3.2Gtps) (256Mbit * 8 @ 128bit) = 51.2GB/s
or
128MB @ 64bit for Cell + 128MB @ 64bit for GPU = 25.6GB/s + 25.6GB/s
while Xenon is
512MB GDDR3 (1.6Gtps) (512Mbit * 8 @ 256bit) = 51.2GB/s
or
256MB GDDR3 (256Mbit * 8 @ 256bit) = 25.6GB/s
Anyway, read the picture above "Optimized XDR core timings yield higher effective bandwidth under random workloads"
PSP has 32MB ram+8MBflash, PS3 128MB?? HUMOUR!
1. 256MB with 25GB/s
2. 512MB with 50GB/s
3. 1024MB with 100GB/s
PSP has 32MB ram+8MBflash, PS3 128MB?? HUMOUR!
PSP = $185 including LCD
PS3 = $? including Cell, media processor, Blu-ray drive, whatsoever
1. 256MB with 25GB/s
2. 512MB with 50GB/s
3. 1024MB with 100GB/s
What are those numbers? If you assume 512MB XDR DRAM, the total bandwidth are 100GB/s.
version
15-Dec-2004, 16:16
PSP has 32MB ram+8MBflash, PS3 128MB?? HUMOUR!
PSP = $185 including LCD
PS3 = $? including Cell, media processor, Blu-ray drive, whatsoever
1. 256MB with 25GB/s
2. 512MB with 50GB/s
3. 1024MB with 100GB/s
What are those numbers? If you assume 512MB XDR DRAM, the total bandwidth are 100GB/s.
CELL dont write too much to memory in game only read that , gpu too
25 GB/s / 60 frame = 416MB/frame , this is enough for 256 MB ram
Brimstone
15-Dec-2004, 17:53
Arriving at final conclusions about the X-Box 2 memory architechture and it's performance might be a little pre-mature at this point. I for one won't be surprised to see something special because of ATI and their history of spearheading the technology assualt on memory issues. The xddr rumor involving VIA and ATI in paticular comes to mind. It will be intresting to see if something sophisticated ends up in the next X-Box.
london-boy
15-Dec-2004, 17:59
1. 256MB with 25GB/s
2. 512MB with 50GB/s
3. 1024MB with 100GB/s
Why do you keep scaling memory size proportionally with Bandwidth?
Why do you keep scaling memory size proportionally with Bandwidth?
bandwith x memory module.
I believe One is right ...Version is off mark by a factor 2
version
15-Dec-2004, 18:07
1. 256MB with 25GB/s
2. 512MB with 50GB/s
3. 1024MB with 100GB/s
Why do you keep scaling memory size proportionally with Bandwidth?
cpu+gpu do read about full memory in every frame at once(geometry,texture)
if memory faster this is wastage
version
15-Dec-2004, 18:31
ps2 32MB - 3.2GB/s
xbox 64MB - 6.4GB/s
128- 12
256-24 .......
this is consol's balance of LAW :)
Fafalada
15-Dec-2004, 19:07
Why do some people have a problem that XeCPU is a clean design, designed and built from scrarch for this one job.
Because it's so much easier to delude oneself that it's a 3.5Ghz Power5 with 3 cores and thus supercomputer equivalent? :P
version
15-Dec-2004, 19:18
Why do some people have a problem that XeCPU is a clean design, designed and built from scrarch for this one job.
Because it's so much easier to delude oneself that it's a 3.5Ghz Power5 with 3 cores and thus supercomputer equivalent? :P
xbox equ with PC
xenon with MAC :D
Why do some people have a problem that XeCPU is a clean design, designed and built from scrarch for this one job.
Because it's so much easier to delude oneself that it's a 3.5Ghz Power5 with 3 cores and thus supercomputer equivalent? :P
That and MS hasn't done enough declaring it to be a huge leap in computing architecture.
People believe what they want to believe even when there are real facts to dispute their beliefs, so in this case where there is no solid info it's hardly surprising.
Why do some people have a problem that XeCPU is a clean design, designed and built from scrarch for this one job.
Because it's so much easier to delude oneself that it's a 3.5Ghz Power5 with 3 cores and thus supercomputer equivalent? :P
That and MS hasn't done enough declaring it to be a huge leap in computing architecture.
People believe what they want to believe even when there are real facts to dispute their beliefs, so in this case where there is no solid info it's hardly surprising.
What would you consider the gecko in gamecube?
That and MS hasn't done enough declaring it to be a huge leap in computing architecture.
MS should declare specs and features not how 'a huge leap is their CPU'.
Is it public how many moneys MS invested in Xenon CPU development?
Tell me how much it costs..and I tell you how fast it is.. :)
ciao,
Marco
Don't forget how long its been under development it is another gauge on just how custom it is. :D
Megadrive1988
15-Dec-2004, 22:36
Li Mu Bai wrote:
Why is it that every patent filed by Toshiba/Sony/IBM somehow relates directly to the PS3? Especially when Cell has much further reaching technical implications for all parties involved? Reyes has been Pana's love child since any Cell patent was 1st filed, or the technology discussed. We all know that this amalgamation of companies didn't join together to solely create the PS3. I realize there's some merit in well-grounded intelligent speculation, but when the PS3 spec. sheet is finally revealed I can already foresee disappointment due to this type of thing. It's already been proven that the PS3 is not the behemoth that it was initially touted to be, but a beast nonetheless. Also, how many tech. patents are existing out there that have yet to come to fruition? (engineers will face headaches even in initial Reyes implementation, not to mention the external memory requirements for essentially a stream processor, ala the PS3 & its subpar random memory access for this type of full application.) Reyes has some time yet before it arrives & creates a paradigm shift in console or PC GPUs. (for many of the developers are one in the same, & this generation will see even more PC dev. cross-platform work, due to the more powerful console architectures & advanced feature sets identical to those of the high-end cards. Surpassing them even for a time) PS4, XBX2, & the NES6 is far more feasible for full Reyes utilization.
It will indeed be some time before the x86 domination of PCs comes to an end as well, unless you're underestimating the time frame in which MS wants/dictates the direction to advance or change. Unfortunately, they also possess the power & consumer mindshare currently to marginalize almost any OS application. Even if it is more efficient, etc. Gasp, I'm tired of reading these Reyes-centric threads. I can still remember when all of those early adherents of Sony's 1 teraflop power claims were defending its technical validity in a multitude of threads. Rolling Eyes Let's wait for more tangible evidence that this indeed is a real possibility, & not simply the first intersection rays if that.
I tend to agree with some of the things you've said here. although other things you said, I don't understand. but lets focus on where we agree.
I believe that alot of the things that some people were hoping for on PS3 will not happen until PS4. like fully Reyes rendering. raytracing. GI. etc.
It's already been proven that the PS3 is not the behemoth that it was initially touted to be, but a beast nonetheless.
I agree that will probably end be being true.
some of the patents filed will probably not be implemented until PS4. I don't think we'll see much if any raytracing until PS4. maybe not even on Xbox3 if X3 comes out by 2009-2010. I don't think we will see even moderate amounts of medium quality raytracing on complex objects and in evironments until well into the next decade. even that might be too soon. but i would hope not. I'd like to see this stuff before i'm too old & grey to see videogames 8)
PS3 will indeed be a beast. but not a GodStation (tm) or RayStation (tm)
:lol:
Fafalada
15-Dec-2004, 22:52
What would you consider the gecko in gamecube?
There's not much to consider - Gekko is a customized version of PPC750cx core. The docs are crystal clear about it, just as they are about PSP CPU being a modified R4000.
(it's actually instruction level compatible(backwards) with 7xx series, so in theory it could be possible to make GCN run older versions of MacOS).
I don't think we will see even moderate amounts of medium quality raytracing on complex objects and in evironments until well into the next decade. even that might be too soon. but i would hope not. I'd like to see this stuff before i'm too old & grey to see videogames 8)
Seeing how relatively well realtime software raytracing works on a boggstandard PC, I would think that, at least shadow raytracing, would be doable nextgen.
What would you consider the gecko in gamecube?
There's not much to consider - Gekko is a customized version of PPC750cx core. The docs are crystal clear about it, just as they are about PSP CPU being a modified R4000.
(it's actually instruction level compatible(backwards) with 7xx series, so in theory it could be possible to make GCN run older versions of MacOS).
So xbox2's cpu won't be compatible with the PowerPC line then? How similar will the final cpu be to the dev kits?
What would you consider the gecko in gamecube?
There's not much to consider - Gekko is a customized version of PPC750cx core. The docs are crystal clear about it, just as they are about PSP CPU being a modified R4000.
(it's actually instruction level compatible(backwards) with 7xx series, so in theory it could be possible to make GCN run older versions of MacOS).
So xbox2's cpu won't be compatible with the PowerPC line then? How similar will the final cpu be to the dev kits?
How did you reach that conclusion based on Faf's statement.
The fact that the CPU core going into the next XBox *might be* a clean design doesn't mean that it doesn't implement the PowerPC ISA.
Considering the dev kits are PPC970 based I think it's fairly certain that the end product will be similar in features. In particular the SIMD extensions (Altivec, VMX, whatever) are well designed (elegant and powerful).
Cheers
Gubbi
What would you consider the gecko in gamecube?
There's not much to consider - Gekko is a customized version of PPC750cx core. The docs are crystal clear about it, just as they are about PSP CPU being a modified R4000.
(it's actually instruction level compatible(backwards) with 7xx series, so in theory it could be possible to make GCN run older versions of MacOS).
So xbox2's cpu won't be compatible with the PowerPC line then? How similar will the final cpu be to the dev kits?
How did you reach that conclusion based on Faf's statement.
The fact that the CPU core going into the next XBox *might be* a clean design doesn't mean that it doesn't implement the PowerPC ISA.
Considering the dev kits are PPC970 based I think it's fairly certain that the end product will be similar in features. In particular the SIMD extensions (Altivec, VMX, whatever) are well designed (elegant and powerful).
Cheers
Gubbi
So its relation to PPC970 might be more like athlon to a 386 than xcpu to pentium 3?
So its relation to PPC970 might be more like athlon to a 386 than xcpu to pentium 3?
Exactly. It'll implement the PPC Instruction Set Architecture (ISA), but might be a completely new core (altivec and all).
Cheers
Gubbi
It's already been proven that the PS3 is not the behemoth that it was initially touted to be, but a beast nonetheless....
...I can still remember when all of those early adherents of Sony's 1 teraflop power claims were defending its technical validity in a multitude of threads. Rolling Eyes
Hate to nit-pick, but this is incorrect. All we know is that the PE concept is valid as many here have been stating and that they've exceeded their clock estimates by a significant margin when fabricated on 90nm sSOI.
Basically, we know that the modular PE is correct; what's still an unknown is how concurrent they can fabricate it on 65nm -- which will ultimately determine the processing capabilities. I wouldn't be surprised if it came out during an ISSCC presentation.
Li Mu Bai
16-Dec-2004, 09:13
Hate to nit-pick, but this is incorrect. All we know is that the PE concept is valid as many here have been stating and that they've exceeded their clock estimates by a significant margin when fabricated on 90nm sSOI.
Basically, we know that the modular PE is correct; what's still an unknown is how concurrent they can fabricate it on 65nm -- which will ultimately determine the processing capabilities. I wouldn't be surprised if it came out during an ISSCC presentation.
I do respect your opinion Vince, but I still don't see the PS3 attaining that kind of computing power. I've conceeded that it will indeed be powerful, though until it actually *acheives* a 10th to the 12th power floating-point operations per second benchmarked, I must remain skeptical. IIRC, these same presentations, papers, etc. had the PS2 capable of accomplishing things no one has seen yet, & no developer has been able to extract. And I suppose that the PS4 will acheive petaflop speeds? :wink:
Considering the dev kits are PPC970 based I think it's fairly certain that the end product will be similar in features. In particular the SIMD extensions (Altivec, VMX, whatever) are well designed (elegant and powerful).
VMX is o.k. but there are better SIMD architectures on the way :wink:
IIRC, these same presentations, papers, etc. had the PS2 capable of accomplishing things no one has seen yet, & no developer has been able to extract.
Did it? Have you seen ISSCC '99 papers digest?
Unfortunately pdf on http://www.isscc.org/isscc/1999/digest/Tuesday.htm are gone now.
Panajev2001a
16-Dec-2004, 13:55
Considering the dev kits are PPC970 based I think it's fairly certain that the end product will be similar in features. In particular the SIMD extensions (Altivec, VMX, whatever) are well designed (elegant and powerful).
VMX is o.k. but there are better SIMD architectures on the way :wink:
APUs/SPUs or the beefed-up VMX units of Xenon/Xbox 2's CPU ;) ?????????????
Brimstone
16-Dec-2004, 14:17
Altivec 2?
Fafalada
16-Dec-2004, 20:19
though until it actually *acheives* a 10th to the 12th power floating-point operations per second benchmarked, operating under real-world gaming complexity scenarios, I must remain skeptical
Programmable FLOPS never refer to some nebulous fantasy of "real-world" usage - it's a simple benchmark of what's the fastest operations the computation units can do.
Any kind of real code will always be lower then that, even with 0 latencies and theoretical infinite bandwith available to the unit.
Because nowadays most vector units are pretty similar (1cycle dotproduct or muladd) directly comparing FLOP number across platforms actually has some meaning - so long as we stick to comparing non hardwired units.
though until it actually *acheives* a 10th to the 12th power floating-point operations per second benchmarked, operating under real-world gaming complexity scenarios, I must remain skeptical
Programmable FLOPS never refer to some nebulous fantasy of "real-world" usage - it's a simple benchmark of what's the fastest operations the computation units can do.
Any kind of real code will always be lower then that, even with 0 latencies and theoretical infinite bandwith available to the unit.
Because nowadays most vector units are pretty similar (1cycle dotproduct or muladd) directly comparing FLOP number across platforms actually has some meaning - so long as we stick to comparing non hardwired units.
Although it doesn't measure the practicalities of feeding the units in various usage scenarios or the difficulties in hiding latency incurred in using the units, or memory.
Considering the dev kits are PPC970 based I think it's fairly certain that the end product will be similar in features. In particular the SIMD extensions (Altivec, VMX, whatever) are well designed (elegant and powerful).
VMX is o.k. but there are better SIMD architectures on the way :wink:
Do you have a concrete example ? Better how ?.
Inquiring minds want to know :)
Cheers
Gubbi
Pure speculation, but how about a superscalar SIMD design
with OOOE and SMT (4 threads say). Drop the PPC FPU, keep the integer units, and add multiple VMX units, allowing for dispatch of multiple SIMD ops out of order.
Fafalada
16-Dec-2004, 23:06
Although it doesn't measure the practicalities of feeding the units in various usage scenarios or the difficulties in hiding latency incurred in using the units, or memory.
Of course, but that would be much harder to clearly present with mere paper specs.
Do you have a concrete example ? Better how ?.
Actually while we're at it, I would argue that PSPs VFPU is already a step forward from VMX (for FPU portion that is, there's no integer SIMD there though :( ).
Do you have a concrete example ? Better how ?.
Actually while we're at it, I would argue that PSPs VFPU is already a step forward from VMX (for FPU portion that is, there's no integer SIMD there though :( ).
Do you have a link to some information on the CPU/VFPU? Googling turns up a sea of press releases but very little info regarding architecture and instruction sets.
Edit: Is it like PS2's VU0/1 ? In that case it's really not an SIMD extension but rather a stand alone SIMD vectur unit, right?
Cheers
Gubbi
Do you have a concrete example ? Better how ?.
Inquiring minds want to know :)
Any details and I'd have to kill you :-(
But taking a general tack, its fairly easy to see how we can improve a SIMD unit for the modern CPU landscape.
Problem A: RAM speeds
A SIMD unit can use alot of RAM (a 4x4 float matrix takes 1/2KB). RISC memory units are too slow (Load/Store as seperate instructions) what we want is old fashioned CISC direct to/from memory. Of course we actually want a small pool of very fast RAM. Lets call that the "register pool", saves any embarassment from RISC fans :-)
So solution to problem A is to have so many registers, its uses the same amount of memory as 8 bit computers used to have. Cell SPU has mentioned 128 128bit registers (16KB), which sounds a good figure.
Problem B: RAM speeds
O.K even with lots of registers I have to read/write stuff sometimes. If I'm going to it would be good to compress everything, say using a decoder like that is fitted to every vertex shader (including PS2) to unpack/pack data.
So solution to problem B is to have dedicated instructions/units for packing/unpacking in the formats most likely to be encountered by GPUs and CPUs.
Problem C: RAM speeds
Still sometimes we are going to stall due to memory latency, so if that happens lets makes sure we have some thread contexts we can switch to see if they could be doing somethin useful.
So solution to problem C is the have multiple thread contexts per core. If one thread stalls, switch to another and do some useful work.
Problem D: We need to pretend that FLOPs figures are really important.
SIMD ALUs are cheap, so lets have a few. Makes the paper figures look good, even though the real problems are A, B and C.
So solution to problem D is to have N SIMD cores.
Note: I'm being overly sarcastic ;-) There are lots of good reasons why having multiple cores is a good things. Its just finding more than about 2 non graphical math intensive (physics and sound are the obvious candidates) tasks gets real hard quickly.
A good SIMD unit will address at least 2 of theses, a really good one will address all 4... The last two are really CPU architecture issues, but the SIMD units have to integrated into the thread architecture to get good performance.
well..AFAIK CELL tries to solve at least a couple of problems (A and D)
I'd add another problem:
E: special scalar units count and latency
Most of the time on SIMD units scalar operations such as reciprocal or square root are deadly slow (much longer latency than fmadd ops and not fully pipelined). That could be mitigated in the same way DeanoC proposes to solve problem C. Another solution is to have more RCP/SQRT units per SIMD unit (see PS2 VU1).
ciao,
Marco
so if I said
1. multiple SIMD ALUs per core
2. 4-way double precision SIMD
Which lines would you put smiley faces next to?
Do you have a concrete example ? Better how ?.
Inquiring minds want to know :)
Any details and I'd have to kill you :-(
But taking a general tack, its fairly easy to see how we can improve a SIMD unit for the modern CPU landscape.
Problem A: RAM speeds
A SIMD unit can use alot of RAM (a 4x4 float matrix takes 1/2KB). RISC memory units are too slow (Load/Store as seperate instructions) what we want is old fashioned CISC direct to/from memory. Of course we actually want a small pool of very fast RAM. Lets call that the "register pool", saves any embarassment from RISC fans :-)
So solution to problem A is to have so many registers, its uses the same amount of memory as 8 bit computers used to have. Cell SPU has mentioned 128 128bit registers (16KB), which sounds a good figure.
Ok, so you want more registers, see below why that might not be *that* good an idea. And you do want to keep values in registers, go to memory and the latency will kill you; even with superfast SRAM you're looking at 3 cycles for both loads and stores (6 i total). Compared to one or two cycles for a result-forwarding network.
Problem B: RAM speeds
O.K even with lots of registers I have to read/write stuff sometimes. If I'm going to it would be good to compress everything, say using a decoder like that is fitted to every vertex shader (including PS2) to unpack/pack data.
So solution to problem B is to have dedicated instructions/units for packing/unpacking in the formats most likely to be encountered by GPUs and CPUs.
A valid point. In altivec you'd have to pack/unpack in software, using a few instructions.
Problem C: RAM speeds
Still sometimes we are going to stall due to memory latency, so if that happens lets makes sure we have some thread contexts we can switch to see if they could be doing somethin useful.
So solution to problem C is the have multiple thread contexts per core. If one thread stalls, switch to another and do some useful work.
I agree, multple thread contexts could boost throughput. But IMO this clashes with your request in A. Each 128 register context would be 2KB. Supporting a second context will be prohibitively expensive. I'd wager you'd be better off with 4 32 register contexts and 64 rename registers and then still have die area to spare.
Problem D: We need to pretend that FLOPs figures are really important.
SIMD ALUs are cheap
<snip>
Its just finding more than about 2 non graphical math intensive (physics and sound are the obvious candidates) tasks gets real hard quickly.
But you'd want to dice each of these into chunk that can be processed in parallel by all your cores, right ?
Cheers
Gubbi
so if I said
1. multiple SIMD ALUs per core
2. 4-way double precision SIMD
Which lines would you put smiley faces next to?
1. The problem is keeping a single SIMD ALU per core busy, so adding another one might not be the best way to spend those transistors.
2. Doubles aren't ready for very high performance SIMD yet IMHO (4 way increases per register to 256bits...).
I agree, multple thread contexts could boost throughput. But IMO this clashes with your request in A. Each 128 register context would be 2KB. Supporting a second context will be prohibitively expensive. I'd wager you'd be better off with 4 32 register contexts and 64 rename registers and then still have die area to spare.
Register renaming doesn't help with large data sets, SIMD often have large sets of matrices (think about skinning or solving a large LCP).
You need direct access to a fairly large memory space, renaming doesn't help here. Hence having 2K of very fast direct access RAM is a good route.
Another problem is that register renaming (like OOOE) isn't something console processors do.
But you'd want to dice each of these into chunk that can be processed in parallel by all your cores, right ?
Its not the easy in practise, I agree its the right way but there is a lot of inertia toward large threads rather than task based chunks.
What about the addition of an L1 D-cache dedicated to the SIMD unit? I imagine it would be able to support reading and writing 128bits per cycle... Support for non-allocating memory operations would reduce cache thrashing for streaming data.
Most of the time on SIMD units scalar operations such as reciprocal or square root are deadly slow (much longer latency than fmadd ops and not fully pipelined).
Eh, RCPSS has a 3 cycle latency on K7/K8, that's pretty fast IMO. The SIMD version RCPPS has 4 cycles latency. Use Newton-Raphson (which is fully pipelined with FMADD) to increase precision if you need it.
Cheers
Gubbi
What about the addition of an L1 D-cache dedicated to the SIMD unit? I imagine it would be able to support reading and writing 128bits per cycle... Support for non-allocating memory operations would reduce cache thrashing for streaming data.
like this?
A load access pattern of the DMA mechanism is detected. At least one potential load of data is predicted based on the load access pattern. In response to the prediction, the data is prefetched from a system memory to a cache before a DMA command requests the data. (http://tinyurl.com/4cgge)
it's about a prefetch mechanism coupled with a L1 cache that feeds a SIMD unit. (CELL stuff..of course ;) )
ciao,
Marco
Eh, RCPSS has a 3 cycle latency on K7/K8, that's pretty fast IMO. The SIMD version RCPPS has 4 cycles latency.
Wow! that's fast :) that unit should be huge ;)
nv20 vertex shaders rcp units have 6 cycles latency AFAIK, PS2 VU1 has a 7 cycles latency div unit and a 11 cycles latency reciprocal unit IIRC.
ciao,
Marco
Register renaming doesn't help with large data sets, SIMD often have large sets of matrices (think about skinning or solving a large LCP).
You need direct access to a fairly large memory space, renaming doesn't help here. Hence having 2K of very fast direct access RAM is a good route.
Completely correct, you of course need enough registers to hold your problem's data. How many matrices are used in skinning these days?
As for solving LCPs you'd probably never have enough registers for that (arbitrary large data set).
Another problem is that register renaming (like OOOE) isn't something console processors do.
Both XCPU and Gecko has register renaming.
Cheers
Gubbi
Eh, RCPSS has a 3 cycle latency on K7/K8, that's pretty fast IMO. The SIMD version RCPPS has 4 cycles latency.
Wow! that's fast :) that unit should be huge ;)
nv20 vertex shaders rcp units have 6 cycles latency AFAIK, PS2 VU1 has a 7 cycles latency div unit and a 11 cycles latency reciprocal unit IIRC.
Not big at all. It's implemented using a small lookup table. And you need to do one or two NR interations on the data to get (almost) full precision. So latency would be comparable to the others, but pipelined (much higher throughput).
Cheers
Gubbi
nAo -
Similar, but thats all about the prefetch mechanism (to the cache). What I propose is more like this :
_____ __________ __________ ____________
| | | | | | |
| | L2 cache | <-> | L1 cache | <-> | SIMD Units |
Mem | <-> | | (1) | | (1) | Load/Store |
| | | ---------- | |
| | | | |
| | | <------------------> | |
| ---------- (2) | |
----- ------------
| __________ |
<-------------------- | Prefetch | <------------
| (3) | (addresses)
----------
(1) standard load/store
(2) non-allocating load/store
(3) Prefetch hardware (prefetch to L2 or L1),
with HW speculation as well as software
instructions
The L2 makes this less relevant to Cell and more so to my imaginary XeCPU. I imagine the L2 cache to be organized into 4 banks
so that 4 cores could load/store to it simultaneously in the absence of bank conflicts.
version
17-Dec-2004, 01:52
if gpu push 3 Gpoly/s, then in a frame push 50 mpoly
1 triangle about 1-10 pixels size
50-500 megapixel per frame with pixelshader , this is too much
with defferred rendering must have only 1 megapixel per frame compute with shader
this technic is 50-500 times faster !!!!!!!!!
nvidia will use it?????
if gpu push 3 Gpoly/s, then in a frame push 50 mpoly
1 triangle about 1-10 pixels size
50-500 megapixel per frame with pixelshader , this is too much
with defferred rendering must have only 1 megapixel per frame compute with shader
this technic is 50-500 times faster !!!!!!!!!
nvidia will use it?????
Don't you think your overdraw is a little high. Optimally at most their should really be only 2x or 3x overdraw on average unless the game engine isn't using any sorting of some sort.
version
17-Dec-2004, 02:21
if gpu push 3 Gpoly/s, then in a frame push 50 mpoly
1 triangle about 1-10 pixels size
50-500 megapixel per frame with pixelshader , this is too much
with defferred rendering must have only 1 megapixel per frame compute with shader
this technic is 50-500 times faster !!!!!!!!!
nvidia will use it?????
Don't you think your overdraw is a little high. Optimally at most their should really be only 2x or 3x overdraw on average unless the game engine isn't using any sorting of some sort.
ok, if to continue your idea, then gpu compute 50-500 Z compare and 2-3 mils shadercycle
with deferred rendering compute 50-500 Z, and 1 mill shader
my conclusion : gpu must have more simple rasterizng unit for Z compare and less pixelshader pipeline
Do you have a concrete example ? Better how ?.
Inquiring minds want to know :)
Any details and I'd have to kill you :-(
But taking a general tack, its fairly easy to see how we can improve a SIMD unit for the modern CPU landscape.
Problem A: RAM speeds
A SIMD unit can use alot of RAM (a 4x4 float matrix takes 1/2KB). RISC memory units are too slow (Load/Store as seperate instructions) what we want is old fashioned CISC direct to/from memory. Of course we actually want a small pool of very fast RAM. Lets call that the "register pool", saves any embarassment from RISC fans :-)
So solution to problem A is to have so many registers, its uses the same amount of memory as 8 bit computers used to have. Cell SPU has mentioned 128 128bit registers (16KB), which sounds a good figure.
Problem B: RAM speeds
O.K even with lots of registers I have to read/write stuff sometimes. If I'm going to it would be good to compress everything, say using a decoder like that is fitted to every vertex shader (including PS2) to unpack/pack data.
So solution to problem B is to have dedicated instructions/units for packing/unpacking in the formats most likely to be encountered by GPUs and CPUs.
Problem C: RAM speeds
Still sometimes we are going to stall due to memory latency, so if that happens lets makes sure we have some thread contexts we can switch to see if they could be doing somethin useful.
So solution to problem C is the have multiple thread contexts per core. If one thread stalls, switch to another and do some useful work.
Problem D: We need to pretend that FLOPs figures are really important.
SIMD ALUs are cheap, so lets have a few. Makes the paper figures look good, even though the real problems are A, B and C.
So solution to problem D is to have N SIMD cores.
Note: I'm being overly sarcastic ;-) There are lots of good reasons why having multiple cores is a good things. Its just finding more than about 2 non graphical math intensive (physics and sound are the obvious candidates) tasks gets real hard quickly.
A good SIMD unit will address at least 2 of theses, a really good one will address all 4... The last two are really CPU architecture issues, but the SIMD units have to integrated into the thread architecture to get good performance.
Been spending much time on Xenon lately?
ok, if to continue your idea, then gpu compute 50-500 Z compare and 2-3 mils shadercycle
with deferred rendering compute 50-500 Z, and 1 mill shader
my conclusion : gpu must have more simple rasterizng unit for Z compare and less pixelshader pipeline
Z Compares will be no where on that order if they use any sort of heirarchal Z Buffer. And Nvidia definately has their own heirarchal Z Buffer and call it Z Occlussion Culling. This allows all the way up to the size of a processing block for each pixel pipeline which on recent Nvidia cards are 32x32 allowing you to toss out then 32x32=1024 pixels at once with one compare if they are all hidden and help throw out other pixels without wasting Z-Buffer bandwidth if they aren't all hidden (then who knows exactly how their table breaks down after that). Then again it might not gain Z-Buffer bandwidth since would surprise me if the entire 32x32 blocks are loaded at once but besides Nvidia doubt anyone knows exactly how they are doing it.
Edit: Thinking about not sure how much my idea holds true for small polygons. It works great for larger polygons and heirarchal Z-Buffers do help reduce Z-Bandwidth and if polygons are sent in meshes you prolly can gain similar performance to having one larger polygon.
version
17-Dec-2004, 03:05
ok, if to continue your idea, then gpu compute 50-500 Z compare and 2-3 mils shadercycle
with deferred rendering compute 50-500 Z, and 1 mill shader
my conclusion : gpu must have more simple rasterizng unit for Z compare and less pixelshader pipeline
Z Compares will be no where on that order if they use any sort of heirarchal Z Buffer. And Nvidia definately has their own heirarchal Z Buffer and call it Z Occlussion Culling. This allows all the way up to the size of a processing block for each pixel pipeline which on recent Nvidia cards are 32x32 allowing you to toss out then 32x32=1024 pixels at once with one compare if they are all hidden and help throw out other pixels without wasting Z-Buffer bandwidth if they aren't all hidden (then who knows exactly how their table breaks down after that). Then again it might not gain Z-Buffer bandwidth since would surprise me if the entire 32x32 blocks are loaded at once but besides Nvidia doubt anyone knows exactly how they are doing it.
Edit: Thinking about not sure how much my idea holds true for small polygons. It works great for larger polygons and heirarchal Z-Buffers do help reduce Z-Bandwidth and if polygons are sent in meshes you prolly can gain similar performance to having one larger polygon.
hierarchical Z Buffer not so good for lots of small polygons
Do you have a concrete example ? Better how ?.
Inquiring minds want to know :)
Any details and I'd have to kill you :-(
But taking a general tack, its fairly easy to see how we can improve a SIMD unit for the modern CPU landscape.
(...)
A good SIMD unit will address at least 2 of theses, a really good one will address all 4... The last two are really CPU architecture issues, but the SIMD units have to integrated into the thread architecture to get good performance.
Been spending much time on Xenon lately?
Why do you think that :?: :lol: :wink:
hierarchical Z Buffer not so good for lots of small polygons
Yeah, thats what my edit was about but if programmers use poly meshes its possible to process whole parts of the mesh like a larger polygon to a certain degree. More complex of course but allows for gains elsewhere as most optimizations do.
hierarchical Z Buffer not so good for lots of small polygons
Yeah, thats what my edit was about but if programmers use poly meshes its possible to process whole parts of the mesh like a larger polygon to a certain degree. More complex of course but allows for gains elsewhere as most optimizations do.
What about irregular z-buffer for GPU?
http://www.cs.utexas.edu/users/billmark/papers/ism-TR/TR04-09.pdf
Fafalada
17-Dec-2004, 03:23
What about irregular z-buffer for GPU?
I'd say that would probably make the GPU slower :P (when using the I-Z).
Though I wouldn't mind if NVidia splashed out on one. :oops:
What about irregular z-buffer for GPU?
I'd say that would probably make the GPU slower :P (when using the I-Z).
Though I wouldn't mind if NVidia splashed out on one. :oops:
Is an irregular z-buffer and TBDR mutually exclusive?
version
17-Dec-2004, 03:34
hierarchical Z Buffer not so good for lots of small polygons
Yeah, thats what my edit was about but if programmers use poly meshes its possible to process whole parts of the mesh like a larger polygon to a certain degree. More complex of course but allows for gains elsewhere as most optimizations do.
ok,
if gpu run on 1 GHZ, must have 2-20 z compare per cycle
average: 4 or 8
and 8,16.. pixelpipeline
thinkable 4 units each with 4 pixelpipelines, cell patent will real ?
PC-Engine
17-Dec-2004, 09:54
A sneek peek at PS3 PC GUI and OS.
http://info.linspire.com/mailers/michaelsminute/video-big.gif
http://images.linspire.com/losshot-big.gif
london-boy
17-Dec-2004, 09:59
A sneek peek at PS3 PC GUI and OS.
http://info.linspire.com/mailers/michaelsminute/video-big.gif
http://images.linspire.com/losshot-big.gif
Huh? what's that?
PC-Engine
17-Dec-2004, 10:10
Linspire...or Linux with a GUI.
london-boy
17-Dec-2004, 10:15
Linspire...or Linux with a GUI.
Oh... kinda cute.
PC-Engine
17-Dec-2004, 10:21
Linspire...or Linux with a GUI.
Oh... kinda cute.
There are some PCs being sold right now that actually have Linspire preinstalled. I was thinking about buying Linspire just to try it out on one of my old PCs which I only use for doing light stuff. It was called Lindows before MS sued them.
http://www.linspire.com/
rabidrabbit
17-Dec-2004, 10:27
So is Linspire the OS that is pre-installed on the "Cell" Workstations?
london-boy
17-Dec-2004, 10:29
So is Linspire the OS that is pre-installed on the "Cell" Workstations?
I guess not.
But it's Linux based, so PCE is kinda saying, hey if it's Linux based it might look like this. Which is cool, i like that layout.
rabidrabbit
17-Dec-2004, 10:38
well, there are several other GUI's for Linux.
PC-Engine
17-Dec-2004, 11:05
well, there are several other GUI's for Linux.
True but this GUI is built-in right out of the box. You can go into many computer stores over here and buy a retail boxed Linspire OS like you could Windows and have it up and running on your PC in 10 minutes with GUI. Linspire was design with ease of use in mind for desktops as opposed to servers.
rabidrabbit
17-Dec-2004, 11:16
If Linspire was developed for desktops not servers, what has it got to do with "Cell", "Cell" Workstations and PS3???
PS3 will not have an OS/GUI that has so many functions as Linspire, it'll propably have a simple interface like that "cross media bar" in PSP and PSX.
PC-Engine
17-Dec-2004, 11:21
If Linspire was developed for desktops not servers, what has it got to do with "Cell", "Cell" Workstations and PS3???
PS3 will not have an OS/GUI that has so many functions as Linspire, it'll propably have a simple interface like that "cross media bar" in PSP and PSX.
It was posted to show that a CELL based PC might have an OS like Linspire since we were talking about the possbility of PS3 being used/sold as a PC alternative back on page 2 or 3.
gleemax
18-Dec-2004, 00:32
Judging by that screenshot, Linspire uses KDE (the K Desktop Environment). I've used it on FreeBSD. It's easy to use after getting it running (which can be hard, if you don't know what you're doing like me). More screenshots: http://kde.org/screenshots/kde330shots.php
Could a desktop PS3 use KDE? Sure. You can install KDE on a PS2 (with the Linux kit).
What workstations, you mean the ones that don't really exsist yet? ;)
You can install KDE on a PS2 (with the Linux kit).
Heh. You can install it, sure... wouldn't want to run it, with only 32mb of ram, though...
You can install KDE on a PS2 (with the Linux kit).
Heh. You can install it, sure... wouldn't want to run it, with only 32mb of ram, though...
Yeah, but thats the beauty of having a x11 server, you can run any window-manager, that is fitting your needs best. KDE is just one of many, there is a lot choice out there. A nice overview can be found here (http://xwinman.org/). My fav is BlackBox, lean and very fast.
I don't think this recent answer by NV's David Kirk last month regarding Ray-Casting has been posted yet,
Ray-Casting
Reader Question: Will we see non-uniform rasterization, streaming ray-casts, or equivalent features to enable the kind of graphics we really want—real-time, dynamic lighting with a large number of lights?—Jason_Watkins
David Kirk: Yes.
Over time, everything that has been hardwired and fixed-function will become general-purpose. This will enable much more variety in graphics algorithms and ultimately, much more realism.
The good news, for my job security, is that graphics is still ultimately very very hard. Tracing streaming rays in all directions, reflecting between all of the objects, lights, and fog molecules in parallel is extremely hard. Nature does it . . . in real time! However, nature does it by using all of the atoms in the universe to complete the computation. That is not available to us in a modern GPU :-).
Graphics will continue to be a collection of clever tricks, to do just enough work to calculate visibility, lighting, shadows, and even motion and physics, without resorting to brutishly calculating every detail. So, I think that there's a great future both in more powerful and flexible GPUs as well as ever more clever graphics algorithms and approaches. Continued...
http://www.extremetech.com/article2/0,1558,1744811,00.asp
Irregular, non-uniform rasterization for PS3's NV5x perhaps :?:
You're not getting that for PC or PS3 in that time frame I'm certain.
london-boy
17-Jan-2005, 13:07
Maybe the PS9's NV12x.
Actually, if we consider NVIDIA making a new NV number every 12 months, and an average life cycle of 5 year per playstation, by the time we get to PS9 in 2048, we will have the NV51X, or the GF54xxx. :wink:
Ray Tracing on a Stream Processor, Stanford University, March 2004 (http://graphics.stanford.edu/papers/tpurcell_thesis/)
An interesting, easy to read paper and pertinent since both CPU+GPU will be stream processors in PS3. ;)
You guys *really * need to get this idea out of your head. it's simply not very realistic. PS3 will be powerfull, but not "that" powerful.
Joshua Luna
18-Jan-2005, 06:18
Ray Tracing on a Stream Processor, Stanford University, March 2004 (http://graphics.stanford.edu/papers/tpurcell_thesis/)
An interesting, easy to read paper and pertinent since both CPU+GPU will be stream processors in PS3. ;)
Hi Jaws! :)
I know GPUs are very parallel devices, are they considered stream processors? If so I learned something today!
If not, do you think the PS3 nVidia GPU will be redone in a way so it can be classified as a "stream processor"?
Thanks :)
london-boy
18-Jan-2005, 10:30
Ray Tracing on a Stream Processor, Stanford University, March 2004 (http://graphics.stanford.edu/papers/tpurcell_thesis/)
An interesting, easy to read paper and pertinent since both CPU+GPU will be stream processors in PS3. ;)
Hi Jaws! :)
I know GPUs are very parallel devices, are they considered stream processors? If so I learned something today!
If not, do you think the PS3 nVidia GPU will be redone in a way so it can be classified as a "stream processor"?
Thanks :)
Today's GPUs are really a mix of many different architectures, and Stream Processor fits into one of the descriptions you could find for them, depending on the task.
Ray Tracing on a Stream Processor, Stanford University, March 2004 (http://graphics.stanford.edu/papers/tpurcell_thesis/)
An interesting, easy to read paper and pertinent since both CPU+GPU will be stream processors in PS3. ;)
Hi Jaws! :)
I know GPUs are very parallel devices, are they considered stream processors? If so I learned something today!
...
Modern programmable GPUs are classed as stream processors and so is a CELL processor. I suppose the simplest definition of a stream processor would be one that executes a stream programming language! :P
Both the engineering of the processor and the compiler for said processor would determine how efficiently streaming code will execute and hide latencies. But hand coding can further optimise code.
There's some nice info in my above link on stream processing. It's a phD paper that was funded by NVIDIA with experiments performed on NV3x class hardware. Whether NVIDIA implement some of its recomendations two generations later for NV5x class hardware reamains to be seen for PS3. Though the notion of a CELL stream processor and a NV5x stream processor would need to complement each other. If CELL is a SIMD stream processor then the NV5x could be a MIMD stream processor to complement it perhaps...
Also here's some more info on the Imagine stream processor from Standford University,
http://cva.stanford.edu/imagine/
Simon F
19-Jan-2005, 11:22
Also here's some more info on the Imagine stream processor from Standford University,
http://cva.stanford.edu/imagine/
"contains 21 million transistors and has a die size of 16mm x 16mm"
I'm not really "au fait" with silicon technology (I'm mainly a SW engineer), but that sounds like a reasonably big lump of silicon for only 21million transistors.
PC-Engine
19-Jan-2005, 18:31
Also here's some more info on the Imagine stream processor from Standford University,
http://cva.stanford.edu/imagine/
"contains 21 million transistors and has a die size of 16mm x 16mm"
I'm not really "au fait" with silicon technology (I'm mainly a SW engineer), but that sounds like a reasonably big lump of silicon for only 21million transistors.
Well you'd have to factor in the process technology for a better perspective. What process was it using anyway?
Edit: it's using 150nm process tech so yeah it's pretty big for only 21 million transistors. :shock:
***BUMP***
A recent SCE patent,
Method and apparatus for real-time global illumination incorporating stream processor based hybrid ray tracing (http://appft1.uspto.gov/netacgi/nph-Parser?Sect1=PTO2&Sect2=HITOFF&u=%2Fnetahtml%2FPTO %2Fsearch-adv.html&r=1&f=G&l=50&d=PG01&p=1&S1=20050041024&OS =20050041024&RS=20050041024)
...
39. A computing device, comprising: a graphics processing unit (GPU) capable of determining lighting characteristics for an object in real time, the lighting characteristics defined through a basis function, the GPU including a stream processor configured to split a stream of data associated with the lighting characteristics into multiple simultaneous operations.
40. The computing device of claim 39, wherein the computing device is a video game console.
41. The computing device of claim 39, further comprising: a display screen in communication with the GPU, the display screen configured to present image data representing the object.
42. The computing device of claim 39, wherein the stream processor is a programmable hardware unit capable of executing code that is replicated multiple times.
43. The computing device of claim 42, wherein the code that is replicated multiple times is configured to process one of a ray tracing algorithm and multiply and add operations for data derived from the ray tracing algorithm.
44. The computing device of claim 43, wherein the ray tracing algorithm determines a direct illumination lighting characteristic in real time and the multiply an add operation determine a secondary lighting characteristic in real time.
45. The computing device of claim 39, wherein the GPU is further configured to render the object through a process involving linear interpolation, such that the lighting characteristics are applied to the rendered object.
46. The computing device of claim 39, wherein the basis function is one of a wavelet and a spherical basis function.
...
Stream processor = CELL?
GPU = NV5x?
Video game console = PS3?
-----------
Reference,
Below is SCEA presentation at SIGGRAPH circa 2001 on the future of next gen games,
http://www.research.scea.com/research/pdfs/la-siggraph-panel_2001_DOM.pdf
and
Ray Tracing on a Stream Processor, Stanford University, March 2004 (http://graphics.stanford.edu/papers/tpurcell_thesis/)
Paper funded by nVidia investigating raytracing on stream processors using pixel shaders and Cg.
You just keep praying don't you. :P
You just keep praying don't you. :P
Good lord Sony, give us a 4PE PS3. Make realtime raytracing a good alternative to shader tricks and please lord, give us a whole gig of RAM. Amen. ;)
Fredi
I found that patent a couple of weeks ago (Panajev knows that.. ;) ) and I refused to link it here cause I didn't want to start another PS3/ray tracing discussion. LOL :)
Titanio
17-Mar-2005, 21:21
Plausibility? It's a couple of years old now..it may no longer be relevant, I guess.
Shifty Geezer
17-Mar-2005, 22:24
Why's it not relevant? PS3's been a few years in development during which time they'd invent tech and patent it (or rather, not invent tech but patent it anyway... :roll: )
I must say, the idea of realtime raytracing AND GI is pretty fantastical :D
Spidermate
17-Mar-2005, 22:28
This was posted from Nirey, a respected and trusted poster from PSINEXT,
Hey sorry guys, i have not posted in so long since last time. But anyways i am sorry i couldn't say anything about this sweet deal until it was announced since i was under very tight NDA's. Anyways, now it's official, and i would like to say that this is a co-joint collaboration between sony in-house graphics technology and Nvidia technology. What this means is that this "custom-gpu" is a totally different architecture that is not based on an existing architecture from Nvidia.
More will come soon.
http://psinext.com/forums/viewtopic.php?p=84712#84712
Hmmm...curioser and curioser...REYES?...Raytracing?...GI? :twisted:
Jaws, it goes A LOT deeper than that now. Some of it would probably blow your mind. But I'll keep my lips sealed .
Jaws, it goes A LOT deeper than that now.
Good god, Jaws heart just missed a beat reading that, be careful with such posts. ;)
Fredi
version
17-Mar-2005, 22:44
"Demonstration written by Gabor Nagy, Sony Computer Entertainment America R+D"
:) ps3 with raytracing is TRUE :)
Gabor Nagy homepage (http://www.equinox3d.com/index.html)
http://www.equinox3d.com/resume.html
Spidermate
17-Mar-2005, 22:52
Jaws, it goes A LOT deeper than that now.
Good god, Jaws heart just missed a beat reading that, be careful with such posts. ;)
Fredi
That could mean anything. :wink:
zidane1strife
18-Mar-2005, 01:23
You just keep praying don't you. :P
Good lord Sony, give us a 4PE PS3. Make realtime raytracing a good alternative to shader tricks and please lord, give us a whole gig of RAM. Amen. ;)
Fredi
I second let us pray lad, let us pray :lol:
Plausibility? It's a couple of years old now..it may no longer be relevant, I guess.
Ain't it fall2k3 that's just a little more than a year ago.
Here's something from Gabor Nagy's Curiculum Vitae:
real-time photorealistic rendering (including global illumination with spherical harmonics).
Titanio
18-Mar-2005, 12:41
Ain't it fall2k3 that's just a little more than a year ago.
True enough. I guess it could fit in nicely with when they probably started talking more seriously with NVidia. Still, though..
I found that patent a couple of weeks ago (Panajev knows that.. ;) ) and I refused to link it here cause I didn't want to start another PS3/ray tracing discussion. LOL :)
LOL!
I also remember reading it myself a couple a weeks ago too and I thought you/someone had posted it with the regular weekly patent updates! :P
It wasn't until I saw this thread,
http://www.beyond3d.com/forum/viewtopic.php?t=21258&start=0&postdays=0&postorder =asc&highlight=
...in the 3D forum that reminded me to post it when I realised it wasn't posted anywhere! ;)
This was posted from Nirey, a respected and trusted poster from PSINEXT,
Hey sorry guys, i have not posted in so long since last time. But anyways i am sorry i couldn't say anything about this sweet deal until it was announced since i was under very tight NDA's. Anyways, now it's official, and i would like to say that this is a co-joint collaboration between sony in-house graphics technology and Nvidia technology. What this means is that this "custom-gpu" is a totally different architecture that is not based on an existing architecture from Nvidia.
More will come soon.
http://psinext.com/forums/viewtopic.php?p=84712#84712
Hmmm...curioser and curioser...REYES?...Raytracing?...GI? :twisted:
Jaws, it goes A LOT deeper than that now. Some of it would probably blow your mind. But I'll keep my lips sealed .
Hehe! ;)
Here's something from Gabor Nagy's Curiculum Vitae:
real-time photorealistic rendering (including global illumination with spherical harmonics).
The other inventor, Robin Green has released a paper on spherical harmonics,
http://www.research.scea.com/gdc2003/spherical-harmonic-lighting.html
He's also got a few more papers at SCEA R&D,
http://www.research.scea.com/research/research.html
Shifty Geezer
18-Mar-2005, 14:45
real-time photorealistic rendering (including global illumination with spherical harmonics).
The other inventor, Robin Green has released a paper on spherical harmonics,
http://www.research.scea.com/gdc2003/spherical-harmonic-lighting.html[/quote]Just downloading now. Reading the intro though, I've got ask has anyone put any of this into practise and if not, why not?
Titanio
18-Mar-2005, 15:05
The other inventor, Robin Green has released a paper on spherical harmonics,
http://www.research.scea.com/gdc2003/spherical-harmonic-lighting.html
Just downloading now. Reading the intro though, I've got ask has anyone put any of this into practise and if not, why not?
Spherical Harmonics? I think it's being used in the UE3 engine. I believe there was also a PSP demo featuring it.
Simon F
18-Mar-2005, 15:10
Spherical Harmonics? I think it's being used in the UE3 engine. I believe there was also a PSP demo featuring it.
Judging from the developer forums, it might be easier to count those who aren't using it. :?
vBulletin® v3.8.4, Copyright ©2000-2010, Jelsoft Enterprises Ltd.