XBox 1 Backwards compatibility

[Dave Baumann]

So, now that NV is providing the graphics for Sony, anyone considered the possability of XB1 backwards compatibility being easier for Sony than MS? ;) :lol:

[McFly]

Are you just bored or did you get some infos that point to some strange backward compatibiliy in PS3? :wink:

Fredi

[london-boy]

TROLL!!!11!111!!!11


:twisted:

[Mulciber]

wouldn't there be legal issues there?

[MODEL 3]

Also, it is more possible than your senario, for XBOX2 to have Gamecube compatibility since XBOX2 will have more things in common with gamecube than PS3 with XBOX1 (In fact XBOX2 from a h/w point of view will have possible more things in common with Gamecube than XBOX1, all is left is the software side)

[Mulciber]

Also, it is more possible than your senario, for XBOX2 to have Gamecube compatibility since XBOX2 will have more things in common with gamecube than PS3 with XBOX1 (In fact XBOX2 from a h/w point of view will have possible more things in common with Gamecube than XBOX1, all is left is the software side)

except the media

[Mendel]

So, now that NV is providing the graphics for Sony, anyone considered the possability of XB1 backwards compatibility being easier for Sony than MS? ;) :lol:

Through emulators, certainly.

For example, xbox (1) is able to run Nintendo 64 games perfectly through emulator.

[london-boy]

So, now that NV is providing the graphics for Sony, anyone considered the possability of XB1 backwards compatibility being easier for Sony than MS? ;) :lol:

Through emulators, certainly.

For example, xbox (1) is able to run Nintendo 64 games perfectly through emulator.

Huh? Since when?

[j^aws]

IBM Microelectronics: Focus on Secrecy


IBM Microelectronics is potentially as powerful as Intel, but like a shadow government it often operates behind the scenes. It currently doesn't ship an x86-compatible processor under its own brand, and its motto seems to be, "we are where you least expect us." IBM Micro was instrumental in getting AMD's Athlon 64 processor to market, it builds VIA's new processors, and it was the first to build Transmeta's. It captured the Microsoft Xbox business and now has virtually all of Apple's business. It is also building the processor for the Sony PS3 and the Next Generation Nintendo platform. (It is already in the current Nintendo Cube.) Basically IBM Microelectronics now owns the future of the console-based game market.


It also just announced eFuse, a unique technology that will allow a processor to dynamically alter itself to adjust for changing conditions.


IBM is expected to release an inexpensive x86-compatible processor based on its PowerPC architecture around the time the Xbox 2 is announced next year. This should give IBM Microelectronics the potential to become a nightmare for Intel.


Of the vendors listed, IBM has the only brand that can match Intel's, and it is also the only vendor that can call on resources that might even eclipse Intel's. With one or two acquisitions, IBM could slide into this space. And while it would not be successful against Intel as long as it is part of IBM Corporate, there are indications that it is considering spinning out of IBM.


IBM is in stealth mode, and that could mean it is planning on moving against Intel next year. It could also mean it is perfectly happy to continue to operate under Intel's radar. One lesson we learned this year is that just because someone is capable of having powerful weapons doesn't mean they actually do.

http://www.technewsworld.com/story/Fighting-for-Leadership-in-the-PC-World-37203.html

Doesn't IBM own Cyrix x86 IP and doesn't need to pay royalty to Intel...that would be funny! :lol:

ROFL

[Evil_Cloud]

So, now that NV is providing the graphics for Sony, anyone considered the possability of XB1 backwards compatibility being easier for Sony than MS? ;) :lol:

Through emulators, certainly.

For example, xbox (1) is able to run Nintendo 64 games perfectly through emulator.

Huh? Since when?

Since people hacked the console and put new BIOS and HDD's in it. ;)

But the emulator isn't perfect though, so I've got a far better example:

The Legend of Zelda: OOT running on GameCube with a Nintendo made N64 emulator.

[Guden Oden]

IBM is expected to release an inexpensive x86-compatible processor based on its PowerPC architecture

That article is waay old, and it's as bizarre now as it was then. "x86-compatible" and "powerpc architecture" are pretty much mutually exclusive, especially used in conjunction with "inexpensive". The one who wrote the article you quote must have smoked some serious sh!t, I can't explain it any other way.

As for the topic of the thread, Nvidia may hold ownership of the graphics and I/O-related parts of the hardware, but MS holds the IP for the copyright protection, the OS/dashboard thingy and all that other stuff that makes the xbox tick. Mildly put, I find it pretty unrealistic to believe PS3 will offer XB compatibility just because NV is involved in designing the graphics chip...

[Spidermate]

So, now that NV is providing the graphics for Sony, anyone considered the possability of XB1 backwards compatibility being easier for Sony than MS? ;) :lol:

Have you ever considered the fact that Sony is helping Nvidia develop the GPU while Microsoft had Nivida built their entire GPU? :| Oh wait, you didn't.

[london-boy]

So, now that NV is providing the graphics for Sony, anyone considered the possability of XB1 backwards compatibility being easier for Sony than MS? ;) :lol:

Have you ever considered the fact that Sony is helping Nvidia develop the GPU while Microsoft had Nivida built their entire GPU? :| Oh wait, you didn't.


:shock: The owner of the site made a thread in what he considers his childplay forums... And you take it seriously? :lol:

[Fox5]

So, now that NV is providing the graphics for Sony, anyone considered the possability of XB1 backwards compatibility being easier for Sony than MS? ;) :lol:

Through emulators, certainly.

For example, xbox (1) is able to run Nintendo 64 games perfectly through emulator.

If by perfectly you mean severe graphical glitches in even the simplest n64 games then yes, it can.(I saw screenshots of the xbox emulator made using the sdk and it had severe errors in mario 64, gamecube's emulator made by nintendo is probably better and has been hacked to play other n64 games)

Doesn't IBM own Cyrix x86 IP and doesn't need to pay royalty to Intel...that would be funny!


I thought VIA bought out Cyrix...

[Spidermate]

So, now that NV is providing the graphics for Sony, anyone considered the possability of XB1 backwards compatibility being easier for Sony than MS? ;) :lol:

Have you ever considered the fact that Sony is helping Nvidia develop the GPU while Microsoft had Nivida built their entire GPU? :| Oh wait, you didn't.


:shock: The owner of the site made a thread in what he considers his childplay forums... And you take it seriously? :lol:

I'm confused. Did I say something wrong?

[london-boy]

I'm confused. Did I say something wrong?

Nothing, you just sounded like you actually were gonna try to "prove Dave wrong"... When the whole thread is obviously a joke, just a light hearted thread. :wink:

[Spidermate]

I'm confused. Did I say something wrong?

Nothing, you just sounded like you actually were gonna try to "prove Dave wrong"... When the whole thread is obviously a joke, just a light hearted thread. :wink:

LOL! Ok, I guess that would make some sense if you were to look at it in that way.

[MODEL 3]

Mulciber wrote:
except the media

Well since PS3 XBOX1-compatibility (or XBOX2 Gamecube-compatibility) (if possible) is going to be illegal then the original gamecube game discs which you are refering to is no issue.For example: Although Dreamcast doesn't have a cartridge slot can play SNES games through emulation.

[Dave Baumann]

The statement was fairly glib, however from the hardware perspective its probably more easily achievable for the Sony setup than the MS route. Regardless of whether Sony is "assisting NVIDIA" or not (although, really, what help could they be for the graphics core?) the eventual outcome is likely to be a superset of previous capabilities, which basically leaves the CPU side and that is proably a much more well known art than the graphics side.

It would be the software front that would be somewhat more of an issue though...

[Fox5]

Mulciber wrote:
except the media

Well since PS3 XBOX1-compatibility (or XBOX2 Gamecube-compatibility) (if possible) is going to be illegal then the original gamecube game discs which you are refering to is no issue.For example: Although Dreamcast doesn't have a cartridge slot can play SNES games through emulation.

SNES games are a bit easier to find online than gamecube games are, and I've never seen a mass market drive or adapter to play the original games of 1 system on another.

Besides, just because xbox2 has an ati graphics chip doesn't mean it'll be compatible with gamecube...the r300 architecture was already completely different from flipper, and I doubt the r500 will be any closer. And is a G5 cpu fully compatible with a G3 cpu? Then there's still the problem of compatible software, the xbox basically uses all pc parts but you can't run a pc version of linux on it, and windows won't run on it at all.

[cthellis42]

If CELL is what it's cracked up to be, I rather assumed they'd be emulating the whole shebang in-processor since it's not your typical CPU-to-GPU relationship. (Or failing that, work out a place to stick in an EE+GS, though I rather don't think they want to suck up the extra cost or spread themselves thin again by adding that to the PSTwo/PSP fab-sharing mix.)

Meanwhile, can one of the insiders please answer whether it's more technical feasability or "ownership issues" that stand in the way of the Xbox 2? At one point before, a lot was being made of nVidia-owned procedures that would have to be licensed extra for Microsoft to get real Xbox compatibility in Xenon.

[Dave Baumann]

Meanwhile, can one of the insiders please answer whether it's more technical feasability or "ownership issues" that stand in the way of the Xbox 2?

ERP has mentioned a few times that he thinks there may be some issues with some of the buffer formats that NVIDIA use. One element that does stick out in my mind is the use of DST+PCF (a la 3DMark05); this is a format that is specific to NVIDIA at the moment. I've heard that ATI are trying to tackle that one somehow, but I don't know if thats in relation to XB2.

[Farid]

and I've never seen a mass market drive or adapter to play the original games of 1 system on another.
You forgot about the Master Gear convertor:
http://img47.exs.cx/img47/5065/gearconvertor2ko.jpg

:D

[Farid]

ERP has mentioned a few times that he thinks there may be some issues with some of the buffer formats that NVIDIA use. One element that does stick out in my mind is the use of DST+PCF (a la 3DMark05); this is a format that is specific to NVIDIA at the moment. I've heard that ATI are trying to tackle that one somehow, but I don't know if thats in relation to XB2.
There's a lot of rumors coming from very trustworthy sources stating that MS is indeed working on BC, and since Nvidia do not help them in anyway, they try to do it by themselves (With the help of Ati or not, that i don't know, but it's likely).
If they can have a working version before the system is considered finish they will announce it, if they can't, they'll do what they're doing now, in other words, they'll do without (They never promised BC for a reason).

[V3]

BC is not a big deal, its just useful in consumer mind set. IMO they don't need 100% BC, just BC the popular games.

[Vince]

Regardless of whether Sony is "assisting NVIDIA" or not (although, really, what help could they be for the graphics core?)

You never did answer what the difference in computational constructs are between an APU and an ALU in X2. I'm interested to hear how the unified SIMD|Scalar datapaths are so different.

And then you can elaborate on, how with the ATI GPU which has unified shaders at the front-end leading upto a fixed-functional back-end that does rasterization, what nVidia and STI can each potentially contribute.

[AzBat]

Regardless of whether Sony is "assisting NVIDIA" or not (although, really, what help could they be for the graphics core?)

You never did answer what the difference in computational constructs are between an APU and an ALU in X2. I'm interested to hear how the unified SIMD|Scalar datapaths are so different.

And then you can elaborate on, how with the ATI GPU which has unified shaders at the front-end leading upto a fixed-functional back-end that does rasterization, what nVidia and STI can each potentially contribute.

Come on Vince give it a rest.

Tommy McClain

[[maven]]

BC is not a big deal, its just useful in consumer mind set. IMO they don't need 100% BC, just BC the popular games.
But that just screams "customer support nightmare". If they manage a quota similar to Sony's PS1 / PS2 compatibility, that's fine, but if it's something much lower, it won't work.
The biggest culprit is still the HDD (or a gigabyte of flash).

[aaronspink]

You never did answer what the difference in computational constructs are between an APU and an ALU in X2. I'm interested to hear how the unified SIMD|Scalar datapaths are so different.

First difference 64 to 128 hardware contexts versus 1 per pipe. The pipes designed by ATI are explicitly designed for the type of workloads which they face, which has a large latency component. To overcome this large latency a large number of hardware contexts are architected into the design. An APU is just a freaking vector unit.

And then you can elaborate on, how with the ATI GPU which has unified shaders at the front-end leading upto a fixed-functional back-end that does rasterization, what nVidia and STI can each potentially contribute.

STI? Nothing design wise. All they will probably be contributing is integration into the rest of the asic and the manufacturing.

Aaron Spink
speaking for myself inc.

[Tuttle]

STI? Nothing design wise. All they will probably be contributing is integration into the rest of the asic and the manufacturing.

Aaron Spink
speaking for myself inc.

Uh no.

With NVIDIA's only experience in console hardware being the horrendously overpriced and underperforming xbox GPU, you can be certain that Sony has NVIDIA on a tight leash design-wise.

[Qroach]

With NVIDIA's only experience in console hardware being the horrendously overpriced and underperforming xbox GPU, you can be certain that Sony has NVIDIA on a tight leash design-wise.

:roll:

[Tuttle]

With NVIDIA's only experience in console hardware being the horrendously overpriced and underperforming xbox GPU, you can be certain that Sony has NVIDIA on a tight leash design-wise.

:roll:

That's right, keep rolling your eyes.

It's amazing that a good number of people have learned nothing, absolutely NOTHING, from the past three to fours years.

[Qroach]

No I'm rolling my eyes at personal stabs people like yourself take at faceless corporations. It's rather lame. Then again, when it comes to console hype, it REALLY is amazing that a good number of people have learned nothing, absolutely NOTHING, from the past four to five years.

Sony is going to get technology Nvidia was/is planning for the PC. If that's what you mean by a "tight leash" then wow, impressive.

[darkblu]

Sony is going to get technology Nvidia was/is planning for the PC. If that's what you mean by a "tight leash" then wow, impressive.

i am somewhat reluctant to believe nv can pull another 'xgpu' fiasco, especially at sony.

[Qroach]

how was what nvidia did with the Xbox GPU a fiasco? because they had MS behind an 8 ball and thus command a high price? Yes it was over priced, but Nvidia could leverage that price with all th other components they provided MS. Without Nvidia the xbox wouldn't have happened imo.

I'm not going to get into the "under performing" nonsesne. The PS2 regarding graphics functionality was just as guilty of this as any console. Anyway, Sony is getting technology originally planned for PC IMO and they are licensing technology not buying fabbed chips. I'd really like to know what a "tight" leash means.

[darkblu]

how was what nvidia did with the Xbox GPU a fiasco? because they had MS behind an 8 ball and thus command a high price? Yes it was over priced, but Nvidia could leverage that price with all th other components they provided MS. Without Nvidia the xbox wouldn't have happened imo.

I'm not going to get into the "under performing" nonsesne. The PS2 regarding graphics functionality was just as guilty of this as any console. Anyway, Sony is getting technology originally planned for PC IMO and they are licensing technology not buying fabbed chips. I'd really like to know what a "tight" leash means.

"over priced" is a bit of an understatement. it was way out of the console league re price/performance. and with that UMA design it was as good as 'bolted' on something originally meant to be a console. ..or again, maybe not, i'm not sure anymore. whatever, nv managed to sell a pickup to someone who was (supposedly) shopping for a porsche.

[Fox5]

and I've never seen a mass market drive or adapter to play the original games of 1 system on another.
You forgot about the Master Gear convertor:
http://img47.exs.cx/img47/5065/gearconvertor2ko.jpg

:D

Wasn't there one of those on genesis too? Anyhow, I was aware of it, I just didn't think it was a big seller.

[Qroach]

no suprise there. you can't make a porsche in a short period of time, same goes for consoles.

[Vince]

First difference 64 to 128 hardware contexts versus 1 per pipe.

That doesn't have to do with the actual computation pathway, that has to do with the complex built around it. This is where the distinction between an APU and SPU comes in...

I specifically phrased the question so people wouldn't comment as you did and force a correction. Going with the current SPU knowledge, there's nothing preventing the current 32 limit (IIRC) going higher outside of added area costs. Although, since we are basically unsure of how the hierarchy works in a PE concerning PU allocated workload, we can't say this either at this point.

[Dave Baumann]

So why use that if you can get better performance out of units dedicated to the task that its required to to for similar or smaller die sizes? And, how are you going to circumvent texture latency?

[Vince]

So why use that if you can get better performance out of units dedicated to the task that its required to to for similar or smaller die sizes?

First of all, who said it was added preformance? In our thought experiment, 16 S|APUs at ISSCC clock would have a higher output than your 48 unified ALUs in the X2 at developer clock. Also, vastly greater usable flexibility, unified ISA and computational fabric.

And the same way a GPU does it Dave, that's the point. One day I'll get an actual answer out of you on the difference between a unified Vector|Scalar pathway in an APU and one in an ALU -- and then you can justify all your, frankly, shit comments about how STI would have nothing to add, et infinitum.,

[Dave Baumann]

And the same way a GPU does it Dave, that's the point.

But its not the same way, since it isn't dedicated to a single usage scenario (or, at least, as focused); the range of instructions and capabilities within a graphics the are tuned to the scenario that it is going to be used in most of the time, unlike more "flexible" units. The are alway highly tuned towards working through the biggest areas of issue for that scenario, of which texture sample latency has been mentioned a number of times, which more "flexible" units are not going to be tuned to address.

[Vince]

But its not the same way, since it isn't dedicated to a single usage scenario (or, at least, as focused); the range of instructions and capabilities within a graphics the are tuned to the scenario that it is going to be used in most of the time, unlike more "flexible" units.

Dave, we're all past the "fixed" verse "flexible" mentality. Tell me a new tale, Explain to me how. We have, basically, analogous unified SIMD pathways. We have roughly similar complexes built around them. One is a highly tuned SOI processor that clocks almost 10X as high and offers full flexibility and the other doesn't.

The are alway highly tuned towards working through the biggest areas of issue for that scenario, of which texture sample latency has been mentioned a number of times, which more "flexible" units are not going to be tuned to address.

Which means... nothing. Again, with an SPU complex able to handle 32 concurrent contexts that we know of, what prevents this from being raised to 128? Or, we have a smart DMAC and a full fledged RISC core sitting right there which is tasked with arbitration -- you might not needed it as you would on a GPU.

[Dave Baumann]

-- you might not needed it as you would on a GPU.

And I'm taking a pretty safe bet that the inclusion of NVIDIA in this project is a surefire indication that these are the type of issues that haven't been worked around, or at least not yet (probably because die space isn't an infinite resource just yet).

[Vince]

And I'm taking a pretty safe bet that the inclusion of NVIDIA in this project is a surefire indication that these are the type of issues that haven't been worked around, or at least not yet (probably because die space isn't an infinite resource just yet).

And that type of thinking would be fallacious, Dave.

[Dave Baumann]

On what evidence?

[Vince]

On what evidence?

Exactly! See, you didn't even need to post to know the answer -- Post hoc ergo propter hoc

[ninelven]

erased

[Dave Baumann]

Exactly! See, you didn't even need to post to know the answer -- Post hoc ergo propter hoc

What?

What do you think NVIDIA have invested the majority of their engineering time on for their next architecture? When JHH steps up there and says "we've spent the last 18 months working on out next generation architecture and will form the basis of our Sony work" what do you think Sony are buying? If its not the shader core then they may as well scrap the majority of that engineering time and they can probably have just shopped around to anyone - hell, why not take a bunch of Imageon cores and strap some APU's in there, clearly, thats all thats needed :!:

[Vince]

Exactly! See, you didn't even need to post to know the answer -- Post hoc ergo propter hoc

What?

It means you can't draw a conclusion based on events which aren't known to have a casual connection just based on the sequence they happened in.

What do you think NVIDIA have invested the majority of their engineering time on for their next architecture? When JHH steps up there and says "we've spent the last 18 months working on out next generation architecture and will form the basis of our Sony work" what do you think Sony are buying?

Why is it that if I were to quote Ken Kutargi word for word, literally, like you just did, I'd have 4-5 people on my ass telling me not to buy into the hype and PR.

If its not the shader core then they may as well scrap the majority of that engineering time and they can probably have just shopped around to anyone - hell, why not take a bunch of Imageon cores and strap some APU's in there, clearly, thats all thats needed :!:

We don't know the terms of the contract, nor do we know all of Sony's options. We do know they were working with Toshiba and the bulk of that work was on raster functionality, which is amazing, huh? Once again, we come back to the problem of scaling Cell and the need for a visualization/output block. We then note that Sony, historically, sucks at producing rasterization functionality, as seen in the GS and PSP to some extent -- likely do to IP barriers.

[Dave Baumann]

It means you can't draw a conclusion based on events which aren't known to have a casual connection just based on the sequence they happened in.

Quite correct, and yet you clearly have. Funny that.

Why is it that if I were to quote ken Kutargi word for word, literally, like you just did, I'd have 4-5 people on my ass telling me not to buy into the hype and PR.

I’m not suggesting you take it literally I’m suggesting you try and look at what is occurring here and open up your mind to a slightly different scenario than the one that appears to be quite firmly embedded.

We do know they were working with Toshiba and the bulk of that work was on raster functionality, which is amazing, huh?

“Raster functionality” include pixel shading.

However, just try this – remove anything else from you mind and tell us where you think NVIDIA’s biggest value is right now in 3D graphics hardware knowledge and application...

[Vince]

“Raster functionality” include pixel shading.

For for the X2, "rasterization" includes Vertex shading and the entire pipeline! Awesome!

However, just try this – remove anything else from you mind and tell us where you think NVIDIA’s biggest value is right now in 3D graphics hardware knowledge and application...

Without a doubt, their IP portfolio.

PS. And when I have I made assumptions like you just did based off the order of events?

PPS. So, what about them unified SIMD|Scalar pathways?

[Dave Baumann]

Without a doubt, their IP portfolio.

And I would suggest you think about the primary application and development of that portfolio is now. Hint: Its what they are telling developers will exponentially increase in performance over the coming years.

PS. And when I have I made assumptions like you just did based off the order of events?

I thought it went something along the lines of “Cell BS/VS patent released” --> Vince knows the answers to the computing universe! ;)

PPS. So, what about them unified SIMD|Scalar pathways?

That not going to go anywhere with anyone with the reply “what can’t be done with more die space…”

[Vince]

And I would suggest you think about the primary application and development of that portfolio is now. Hint: Its what they are telling developers will exponentially increase in performance over the coming years.

I did and their IP portfolio doesn't equate to just your fixation on shaders, likely do in no small point because of ATI's influence. Their IP portfolio likely extends to not only their (Sony) work on APU construction and compiler techology, or the fixed-functional aspects of the rasterization, but to data flow and control to software design.

I thought it went something along the lines of “Cell BS/VS patent released” --> Vince knows the answers to the computing universe!

Cute, I was expecting something that happened. But whatever.

That not going to go anywhere with anyone with the reply “what can’t be done with more die space…”

That wasn't my answer, you're avoiding the question. The question is die size invarient and concerns the actual computation pathway. Your question concerning the SPU complex is self-answering as STI came up with the Synergistic Processor for a reason...

[Dave Baumann]

I did and their IP portfolio doesn't equate to just your fixation on shaders, likely do in no small point because of ATI's influence.

Vince, this is the industry's "fixation on shaders". Go and read some recent developer documation from NVIDIA and you'll see it littered everywhere - even down to elements such as steering developers away from the Doom 3 lighting model because it doesn't scale with shaders, they are telling developers to use SLI to test high shader utlisation now for performance 12 months away.

[Vince]

Vince, this is the industry's "fixation on shaders".

Welcome to my argument against you and Joe from over a year ago based on sheer computational importance vis-a-vis lithography ability. Now, with your little diatribe in mind, explain to me the difference between the computational ability of the unified SIMD Vector|Scalar datapath in an APU and in an ALU.

Back to step one after a few hours of you dancing around.

[aaronspink]

Uh no.


With NVIDIA's only experience in console hardware being the horrendously overpriced and underperforming xbox GPU, you can be certain that Sony has NVIDIA on a tight leash design-wise.

Hmm, well lets see. Overpriced? Underperforming? Compared to what? I'd wager no small amount of money that the production costs for the Nvidia GPU and associated memory is cheaper than the Sony GS. Now what MS pays may not be, but that is a different issue.

Performance would have to the the Nvidia GPU.

The point being that you don't need experience in console hardware design to design graphics and experience in console hardware design does not equal experience in graphics hardware design.

The only reason that Nvidia is designing anything for PS3 is because Sony realized that its internal designs would not be sufficient. I would suspect that the leash that Sony has on Nvidia is made out of a wet noodle.


Aaron Spink
speaking for myself inc.

[Qroach]

Speaking of dodging questions vince, and bringing those to just about eveyr discussion you want to participate in. You never did answer my question from an older thread.

Vince wrote:
"They're not. They have a synergistic, additive effect on sales potential because the more Cell you own, theoretically, the more value you derive from each purchase."

How is the "more you own, the more value you derive" going to play out for the average consumer that can't afford to have "all new" appliences or devices, and can only scrounge enough money together to get a new game system once every four years? What exactly is this value they are getting?

[Tuttle]

Uh no.


With NVIDIA's only experience in console hardware being the horrendously overpriced and underperforming xbox GPU, you can be certain that Sony has NVIDIA on a tight leash design-wise.

Hmm, well lets see. Overpriced? Underperforming? Compared to what? I'd wager no small amount of money that the production costs for the Nvidia GPU and associated memory is cheaper than the Sony GS. Now what MS pays may not be, but that is a different issue.

Performance would have to the the Nvidia GPU.

The point being that you don't need experience in console hardware design to design graphics and experience in console hardware design does not equal experience in graphics hardware design.

The only reason that Nvidia is designing anything for PS3 is because Sony realized that its internal designs would not be sufficient. I would suspect that the leash that Sony has on Nvidia is made out of a wet noodle.


Aaron Spink
speaking for myself inc.

You don't actually believe that do you? Do you actually think Sony's GS is more costly than the xbox GPU???

After doing little more than giving MS a big fat expensive peecee video card to bolt on to the xbox, NVIDIA will have hopefully been given quite an education in console GPU desgin over the past two years from Sony.

[aaronspink]

First of all, who said it was added preformance? In our thought experiment, 16 S|APUs at ISSCC clock would have a higher output than your 48 unified ALUs in the X2 at developer clock. Also, vastly greater usable flexibility, unified ISA and computational fabric.

Here's some math:

A texture read will take on the order of 40nS.

@4 Ghz this is 160 cycles.
@.5 Ghz this is 20 cycles.

Assume 1 texture read per 8 shader ops.

ATI R500 48*.5 = 24 Giga Ops per cycle. Don't need to account to wait states because of multiple contexts provided in hardware.

Vince's PS3 on crack 16*4/160 = 400 Mega Ops per cycle.

The only way that a non graphics optimized pipeline will win is if the number of ops per texture read is extremely high or the latency of the texture read is extremely low.



And the same way a GPU does it Dave, that's the point. One day I'll get an actual answer out of you on the difference between a unified Vector|Scalar pathway in an APU and one in an ALU -- and then you can justify all your, frankly, shit comments about how STI would have nothing to add, et infinitum.,

I'll provide you your answer: a large number of hardware contexts optimized to cover the texture fetch latency. A building block in the Sony design is not optimized for this workload because it can't handle the texture fetch latency as well as various other operations (sampling, filtering, early Z reject, etc) that are designed into the GPU.

Aaron Spink
speaking for myself inc.

[Vince]

ATI R500 48*.5 = 24 Giga Ops per cycle. Don't need to account to wait states because of multiple contexts provided in hardware.

Vince's PS3 on crack 16*4/160 = 400 Mega Ops per cycle.

Didn't we already discuss what an SPU is and didn't I mention it's widely believed that it can already handle 32 simultaneous contexts? Not taking into account the work of the main DMAC and PU in arbitration.

I'll provide you your answer: a large number of hardware contexts optimized to cover the texture fetch latency.

As I already stated, this isn't dependent upon the unified SIMD Vector|Scalar datapath output (which is what was being discussed), it's dependent upon the complex you build around it. A Synergistic Processor is just this.

A building block in the Sony design is not optimized for this workload because it can't handle the texture fetch latency as well as various other operations (sampling, filtering, early Z reject, etc) that are designed into the GPU.

I fully agree with you on the latter functions, which is what I believe Sony will utilize, the ROP/Pixel Engine functionality. I do think you're incorrect in your assessment of texture access effeciency.

[aaronspink]

Dave, we're all past the "fixed" verse "flexible" mentality. Tell me a new tale, Explain to me how.

Ok, sure, but I'm not quite sure you are going to understand.


We have, basically, analogous unified SIMD pathways.

This would be incorrect.

We have roughly similar complexes built around them.

Where is the texture filtering? ROBs? Where is the Z reject? The complexes have striking differences.

One is a highly tuned SOI processor that clocks almost 10X as high and offers full flexibility and the other doesn't.

A 10 Ghz 386 still isn't going to out perform a 1 Ghz Alpha or 1.3 Ghz Power4 or 1 Ghz Pentium 3 or a 1 Ghz Athlon. Just cause something has a high clock speed doesn't mean its fast.

Aaron Spink
speaking for myself inc.

[Vince]

This would be incorrect.

Where is the difference?

We have roughly similar complexes built around them.

Where is the texture filtering? ROBs? Where is the Z reject? The complexes have striking differences.

I was refering to the ALUs, which are disjointed from the raster-ops in contemporary GPUs AFAIK. The NV40 is like this concerning the shading complexes/units which are composed of 2 ALUs and other associated units which are physically independent of the ROP/Z/Stencil functionality. I assume the X2's unified shaders will follow in this.

A 10 Ghz 386 still isn't going to out perform a 1 Ghz Alpha or 1.3 Ghz Power4 or 1 Ghz Pentium 3 or a 1 Ghz Athlon. Just cause something has a high clock speed doesn't mean its fast.

When the contructs are similar as they are in the unified SIMD Scalar|Vector processors, it surely applies. As I said concerning the answer that the difference lies in the numbers of units, it's not relevant if the compuitational resources lie along the temporal or spatial axis.

[Fox5]

Dave, we're all past the "fixed" verse "flexible" mentality. Tell me a new tale, Explain to me how.

Ok, sure, but I'm not quite sure you are going to understand.


We have, basically, analogous unified SIMD pathways.

This would be incorrect.

We have roughly similar complexes built around them.

Where is the texture filtering? ROBs? Where is the Z reject? The complexes have striking differences.

One is a highly tuned SOI processor that clocks almost 10X as high and offers full flexibility and the other doesn't.

A 10 Ghz 386 still isn't going to out perform a 1 Ghz Alpha or 1.3 Ghz Power4 or 1 Ghz Pentium 3 or a 1 Ghz Athlon. Just cause something has a high clock speed doesn't mean its fast.

Aaron Spink
speaking for myself inc.

How about a 1.5ghz 386 with 64MB L1 cache and 256MB L2 cache?(btw, why did you do straight out 1ghzes but did 1.3 for the power4...afaik alpha has the fastest mhz per mhz performance, then athlon, then power4, then p3 in last)

[3dcgi]

Uh no.


With NVIDIA's only experience in console hardware being the horrendously overpriced and underperforming xbox GPU, you can be certain that Sony has NVIDIA on a tight leash design-wise.

Hmm, well lets see. Overpriced? Underperforming? Compared to what? I'd wager no small amount of money that the production costs for the Nvidia GPU and associated memory is cheaper than the Sony GS. Now what MS pays may not be, but that is a different issue.

Performance would have to the the Nvidia GPU.

The point being that you don't need experience in console hardware design to design graphics and experience in console hardware design does not equal experience in graphics hardware design.

The only reason that Nvidia is designing anything for PS3 is because Sony realized that its internal designs would not be sufficient. I would suspect that the leash that Sony has on Nvidia is made out of a wet noodle.


Aaron Spink
speaking for myself inc.

You don't actually believe that do you? Do you actually think Sony's GS is more costly than the xbox GPU???

After doing little more than giving MS a big fat expensive peecee video card to bolt on to the xbox, NVIDIA will have hopefully been given quite an education in console GPU desgin over the past two years from Sony.
I don't know how you can fault Nvidia for the Xbox having a PC like design with limited bandwidth. The console design for PS3 is done by Sony as Microsoft chose the console design for Xbox. So if you have a problem with Xbox as a whole blame Microsoft.

In PS3 I expect specs like memory bandwidth to be dictated by Sony. As far as the graphics chip design is concerned I'd bet Sony suggested some features, but for the most part the design is all Nvidia. It seems Vince is thinking Sony will help design the ALUs to run at a faster clock rate. Maybe. I don't think anyone reading Beyond3D, that would talk, can answer Vince's question for sure.

[Tuttle]

I don't know how you can fault Nvidia for the Xbox having a PC like design with limited bandwidth. The console design for PS3 is done by Sony as Microsoft chose the console design for Xbox. So if you have a problem with Xbox as a whole blame Microsoft.

In PS3 I expect specs like memory bandwidth to be dictated by Sony. As far as the graphics chip design is concerned I'd bet Sony suggested some features, but for the most part the design is all Nvidia. It seems Vince is thinking Sony will help design the ALUs to run at a faster clock rate. Maybe. I don't think anyone reading Beyond3D, that would talk, can answer Vince's question for sure.

Coulda Shoulda Woulda. The fact remains. Nvidia has one entry on their console hardware resume, and it is a ugly one.

There is no chance Sony just "suggesting features." None. The PS3 GPU is going to be a Sony design from top to bottom. End of story. The low level details of that design will most likely be Nvidia though.

[3dcgi]

There is no chance Sony just "suggesting features." None. The PS3 GPU is going to be a Sony design from top to bottom. End of story. The low level details of that design will most likely be Nvidia though.
You're correct about the "suggesting features" quote. Of course they will dictate the features they'd like to see if Nvidia's design didn't already support them. I just mistyped that. I don't see how you're so sure Sony is designing the GPU from top to bottom though. I wouldn't believe that unless I heard it straight from Nvidia.

[passerby]

There is no chance Sony just "suggesting features." None. The PS3 GPU is going to be a Sony design from top to bottom. End of story. The low level details of that design will most likely be Nvidia though.
Ok, ok. We can just wait for public details to emerge after the presentation in February - and hopefully before the close of Q4 '04. I mean we've already waited like what? 2 years for some half-obsured Cell details. We can wait another 4 months. :shock: Until then the scoop of this discussion will keep hiting a wall.

[ninelven]

erased

[V3]

In PS3 I expect specs like memory bandwidth to be dictated by Sony. As far as the graphics chip design is concerned I'd bet Sony suggested some features, but for the most part the design is all Nvidia. It seems Vince is thinking Sony will help design the ALUs to run at a faster clock rate. Maybe. I don't think anyone reading Beyond3D, that would talk, can answer Vince's question for sure.

IMO there are few scenarios that can happend,

1. Like Dave suggested, PS3 GPU is NV next generation PC part, known as the NV50 that has been rumoured cancelled.

2. Like the first scenario, but Sony is helping NV to get the clock speed up and power down, with their 90nm fabs, which will be quite mature by the time PS3 GPU needs to go into production.

3. Sony contributes the Synergistic Processors, and NV contributes the rasterizer with their pixel shading technology.

4. Like the third, but NV modifies and restructures the synergistic processor to tune it for graphics work.

Just my two cents.

[3dilettante]

How about a 1.5ghz 386 with 64MB L1 cache and 256MB L2 cache?

When you say 386, do you mean the actual old-timer 386 chips?

Even with the impossibly huge caches, the 386 would lose, and lose badly. A simple register to register add takes 2 cycles on a 386 (memory register adds took a minimum 7 cycles). Rudimentary pipelining could shave off a cycle if no address offsets were needed for the next instruction, meaning if the only thing you wanted to do was only repeated register adds, 1.5 billion adds a second are possible.

On an alpha EV6, even without caches that huge, something so simple as endlessly adding can be dispatched 4 times a clock cycle. So to match a 1 Ghz alpha on insanely pointless code, this 386 would have to go 4 Ghz doing only a little more than nothing.

On any code not trivially easy, the lack of pipelining would slam the single-issue 386 down to an execution rate of a fraction of an instruction per cycle. This is assuming the giant cache removes all memory considerations (it won't, especially not with the ISA exposed to a 386).

Far smaller caches have something higher than 80% of that effectiveness already, and with the die area of cache that huge, you could fit a half dozen alphas with far better effect.

A modern processor, unless you are deliberately telling it to idle, will fail miserably at cutting its execution rate anywhere near that low, and it will have floating point capability to boot.

edit: not to say that superscalar processors are all that great at getting more than 1 instruction executed per clock thanks to memory latency and other hazards. I think someone gave an average of .8 instructions per clock for a K8. If the old chips weren't orders of magnitude worse, modern chips would look much less impressive.

[see colon]

and I've never seen a mass market drive or adapter to play the original games of 1 system on another.

besides the SMS->GG (doesn't really count imho because they were effectivly the same hardware) and SMS->gen adapters already mentioned...

nintendo has released the super gameboy (gb->snes) and gameboy player (gb/gbc/gba->gc)

bleemcast was planned and sort of released (in single game packs) for dc (psx->dc)

3do had a hardware card that allowed you to play 3do games on a pc

apple used to sell pc cards (effectivly a pc on a card to run pc apps on a mac)

there was a few psx emulators comercialy available for pc and mac (bleem!, CVGS)

datel released a product known as the "advance game port" for the gc that plugged into the memory card slot on the GC and played gba games (no gb/gbc compatability)

speaking of datel, and in the "how can this be legal" slot, the action replay max for playstaion 2 offers a slew of features including a "retro game emulator". apparently (a friend has one) you can play genesis games on it pretty well, and i've heard rumors that it supports snes also.
http://www.datel.co.uk/products.asp

action replays sell pretty well, too.

[Vince]

So why use that if you can get better performance out of units dedicated to the task that its required to to for similar or smaller die sizes?

I had forgotten, but was reminded, that your claims of smaller die size in the context of the ALU is incorrect. Currently, on both the NV40, the R3xx derived, and I believe the X2 ALUs utilize seperate Scalar|Vector constructs placed in parallel per ALU 'axis.' Gschwind, Hofstee and Altman's APU is based around a unified SIMD Vector|Scalar construct that appears to be more area effecient and which doesn't have the drawbacks seen when using SIMD subword parallelism via partitioned Scalar constructs.

I believe your entire argument is suspect and untenable, which we'll see more clearly come ISSCC when the entire SPU complex is presented.

[Panajev2001a]

Just a question Vince.

Do you believe that Toshiba's "rumored" GPU for PlayStation 3 was CELL based in any or some way ?

Just asking... ;).

[Dave Baumann]

Currently, on both the NV40, the R3xx derived, and I believe the X2 ALUs utilize seperate Scalar|Vector constructs placed in parallel per ALU 'axis.'

Nope. Only on the VS currently, not for the PS. (PS on NV40 and R300 are vector, but capable of a co-issue split - up to Vec3 and scalar for R300, upto Vec3 and scalar or Vec2 + Vec2 on NV40).

[Vince]

Currently, on both the NV40, the R3xx derived, and I believe the X2 ALUs utilize seperate Scalar|Vector constructs placed in parallel per ALU 'axis.'

Nope. Only on the VS currently, not for the PS. (PS on NV40 and R300 are vector, but capable of a co-issue split - up to Vec3 and scalar for R300, upto Vec3 and scalar or Vec2 + Vec2 on NV40).

Perhaps read the entire post first Dave, maybe then respond to what was being talked about?

[Dave Baumann]

You talked about what you thought was the PS structure Vince, I was correcting your incorrect ideas of that structure.

[Vince]

First of all, according to your own site, R3xx has parallel Vector and Scalar constructs, yes or no? And you totally missed my part about achieving subword parallelism and the drawbacks by partitioning for some bizzare reason...

[Dave Baumann]

First of all, according to your own site, R3xx has parallel Vector and Scalar constructs, yes or no?

As your previous post said the Vertex Shader has a parallel vector and scalar ALU processor capabilites such that a "5D" operation can be achieved in a single cycle per unit, the Pixel Shaders on the other hand do not; they are a "4D" unit only (but have differing co-issue capabilties when up to 4 ops are required in two instructions).

[Vince]

As your previous post said the Vertex Shader has a parallel vector and scalar ALU processor capabilites such that a "5D" operation can be achieved in a single cycle per unit, the Pixel Shaders on the other hand do not; they are a "4D" unit only (but have differing co-issue capabilties when up to 4 ops are required in two instructions).

I do believe it was you making the assumption (which I responded and told you to respond to what was written) and it was you who selectively quoted, leaving out the second half of the post.

[Dave Baumann]

Vince, that was a clarficiation on the architectures mentioned, thats all.

[Megadrive1988]

okay, I have not had time to read this entire thread. just wanted to mention that the cnn money article on Xbox Next PC
(3rd version of Xbox 2) would have or might have Xbox1 bc.

is that something that would be acceptable to you guys, the 3rd and most expensive version of Xbox2, coming in 2006, being the only one that could play Xbox1 games?

[Panajev2001a]

You still did not answer to me Vince ;).

[Xenus]

Do you think he would be arguing with Dave about the S/APUs if he didn't. :lol:

[AzBat]

okay, I have not had time to read this entire thread. just wanted to mention that the cnn money article on Xbox Next PC
(3rd version of Xbox 2) would have or might have Xbox1 bc.

is that something that would be acceptable to you guys, the 3rd and most expensive version of Xbox2, coming in 2006, being the only one that could play Xbox1 games?

I believe it was speculated that the 2nd version(with HD) could have it as well. The purpose of backward compatibility IMHO is to have it with launch. It helps with not having enough games. If BC didn't come till 2006 then I'm not sure that would be favorable. Maybe it would be OK if it was promised for a 2006 launch. However, I can see how Microsoft might like that option especially if they were expecting to continue selling Xbox1 consoles and BC wasn't ready. IMHO either MS does it at launch or not at all.

Tommy McClain

[Megadrive1988]

I believe it was speculated that the 2nd version(with HD) could have it as well. The purpose of backward compatibility IMHO is to have it with launch. It helps with not having enough games. If BC didn't come till 2006 then I'm not sure that would be favorable. Maybe it would be OK if it was promised for a 2006 launch. However, I can see how Microsoft might like that option especially if they were expecting to continue selling Xbox1 consoles and BC wasn't ready. IMHO either MS does it at launch or not at all.

Tommy McClain


good points. I tend to agree.

[wco81]

The rumored 3 configuration of Xenon wouldn't be good for overall marketshare. If the performance differs among the 3, then you end up with games being targeted for the LCD, which would make the people who bought the highest-end model unhappy. And the early adopters who bought the low configuration wouldn't be happy to see what they bought eclipsed so quickly.

Worst-case scenario is that whatever extra capabilities of the higher configurations aren't supported by 3rd parties any more than the PS2 HDD.

[psurge]

Could the different versions be targeted at different HDTV resolutions? I mean that each game would run on any of the versions, but if you want the very high res HDTV formats, then you need the more powerful version. I mean if you have 10k to spend on an 1980x1080p (or whatever it is) HDTV, a $500+ X2 "turbo" is not really that big a deal.

If you went with that kind of model, then you could release the initial version
that can handle TV and HDTV (720p max), then later you scale up the #of
shading units on the GPU to handle the crazier TVs. The higher performance model becomes feasible after a process shrink, and this also gives you some time for HDTVs that support really high res to penetrate the market.

Just an idea,
Serge

[AzBat]

From my reading on the 3 SKUs of the Xbox2 is that the games functionality of all 3 would be identical. Therefore there wouldn't be a LCD problem. Unfortunately that means all games will be designed for one standard resolution. However, I believe the prospect for 720p being the standard resolution is pretty high. With that said, a higher percentage of 1080i games is looking pretty good.

IMHO, I wouldn't want 3 different SKUs with varying game performance. Developers won't take advantage of it. This is a game console after all. Also, I wouldn't be surprised if MS did release 2 different Xbox2. I have a feeling that although the base Xbox2 will not have a hard drive it will come with enough built-in flash memory that it will provide BC support. Whether or not MS can pull off BC support is another story. I guess we will find out eventually.

Tommy McClain

[loekf2]

From my reading on the 3 SKUs of the Xbox2 is that the games functionality of all 3 would be identical. Therefore there wouldn't be a LCD problem. Unfortunately that means all games will be designed for one standard resolution. However, I believe the prospect for 720p being the standard resolution is pretty high. With that said, a higher percentage of 1080i games is looking pretty good.

IMHO, I wouldn't want 3 different SKUs with varying game performance. Developers won't take advantage of it. This is a game console after all. Also, I wouldn't be surprised if MS did release 2 different Xbox2. I have a feeling that although the base Xbox2 will not have a hard drive it will come with enough built-in flash memory that it will provide BC support. Whether or not MS can pull off BC support is another story. I guess we will find out eventually.

Tommy McClain

Wait.. wait.. I've also read the rumours that Microsoft will release 3 versions of the Xenon, one even vaguely looking like a normal PC. I have serious doubts whether Microsoft can pull this off. A non-X86 system running something like Windows XP.. exposed to a X86-based world ?

Unless they can emulate a fast AMD32/64 (on purpose) class CPU in this 4 CPU core so many GHz thing.

Nontheless, a 733 MHz PIII ... maybe it's possible after all.

[Fox5]

and I've never seen a mass market drive or adapter to play the original games of 1 system on another.

besides the SMS->GG (doesn't really count imho because they were effectivly the same hardware) and SMS->gen adapters already mentioned...

nintendo has released the super gameboy (gb->snes) and gameboy player (gb/gbc/gba->gc)

bleemcast was planned and sort of released (in single game packs) for dc (psx->dc)

3do had a hardware card that allowed you to play 3do games on a pc

apple used to sell pc cards (effectivly a pc on a card to run pc apps on a mac)

there was a few psx emulators comercialy available for pc and mac (bleem!, CVGS)

datel released a product known as the "advance game port" for the gc that plugged into the memory card slot on the GC and played gba games (no gb/gbc compatability)

speaking of datel, and in the "how can this be legal" slot, the action replay max for playstaion 2 offers a slew of features including a "retro game emulator". apparently (a friend has one) you can play genesis games on it pretty well, and i've heard rumors that it supports snes also.
http://www.datel.co.uk/products.asp

action replays sell pretty well, too.


Ok, you win, I seriously forgot about most of those, and I've seen and read up on all of them, and even own bleem, bleemcast, and the gameboy player. They don't seem as hackish as the master drive convertors though which looked more like something piraters would use to play backups of games.

[see colon]

the master drive converters for the game gear were basicly just "pin conversions". the hardware for the master system and gamegear is essentialy the same, but the cart size was smaller on the gg. i have a 3rd party sms->gg converter, and it does look pretty ghetto. it gets some pretty funny looks if you play it in public, too. huge-ass carts sticking up perpandicular from the gg, it's pretty goofy looking.

[Pete]

With NVIDIA's only experience in console hardware being the horrendously overpriced and underperforming xbox GPU, you can be certain that Sony has NVIDIA on a tight leash design-wise.Didn't nV do the Saturn video and audio, as well? Or am I misremembering the Diamond Edge 3D and its Virtua Fightery goodness?

This is re: nV's "only experience in console hardware," not "overpriced and underperforming."

[one]

Didn't nV do the Saturn video and audio, as well?

Apparently no.

http://www.nvnews.net/vbulletin/printthread.php?t=29935

no no no, Nvidia never made any chips for the Saturn at all. nor where they going to. the Sega-Nvidia partnership was for a chip for a successor to the Saturn. whatever that sucessor was, it went unmentioned, and is also not the Dreamcast.

the Saturn was in development in 1991-1992, before Nvidia even existed. Saturn uses Hitachi CPUs and SEGA-designed 2D sprite & background graphics processors (VDP1 and VDP2) which are derived from Sega's System32 arcade board of 1991. plus Saturn has a bunch of other processors, from Sega, Yamahi.

there is no Nvidia technology in Saturn whatsoever. Nvidia was a new startup company by the time the Saturn was finished (1994).

Nvidia's first chip, the NV1 of 1995, was used in the Diamond Edge 3D card. Sega and Nvidia had another agreement where Sega Saturn games were ported over to the NV1 chip / Diamond Edge 3D card, and Sega Saturn controllers could be plugged into the card--the Diamond Edge 3D card was a multi media card with 3D, 2D, Audio, etc.

the NV2 chip that Nvidia was working on for Sega used quads, like the NV1, but Sega didn't want quads. Nvidia refused to make a chip that pushed triangles. Sega backed away from Nvidia without any drama at all (unlike the well known Sega-3Dfx dispute)

the NV2 chip was scrapped, although it very well might have been used in the Sega PICO. but there's no evidence of that, though (NV2 in PICO comes from the FiringSquad.com article)

After Sega parted ways with Nvidia sometime in 1995, Sega went on to talks with Lockheed Martin over using Real3D in a console. nothing came of those talks. then Sega went to PowerVR and 3Dfx, ultimately selecting the PowerVR2 based Dural in 1997 which was then named Katana, and then finally, Dreamcast 1998.

[pahcman]

Yes another silly "peeeceee" jab by tuttle. Okay if he try to say Sony PS harware are designed to scale with good in house fabbing, but to wipe away Nvidia strenghs, relegating them to bitplayer.... :roll:

Sony has only 2 entry while Nvidia have helped with 1. Nvidia have sold countless generations of graphics cards. Sure, Xbox is a quick bolted 18 months job yet enough to humble the "experienced".

A translated Watch news, which say Sony was so impressed with Nvidia Xbox "peeeceee" graphics, that they now while working on internal PS3 GPU, open for possible "peeeceee" solution, which we know which route they eventually went. Sony fans here have breath a sigh of relief and excited with Nvidia audio and video solution.
Xbox which loads XGPU and XMCP of NVIDIA appearing 2001 November. When of the lead time is thought, SCEI and NVIDIA are thought that already it had started negotiation when Xbox comes out from. As for this, you can think the possibility Xbox producing effect on SCEI.

Actually, as for the GPU industry authorized personnel who is 2002 beginning of the year spring "as for SCEI it was the schedule which originally develops the graphics of PS3 in inside. But, you have heard that Xbox appearing with that much graphic power gave impact to SCEI. Therefore, as for the graphic tip/chip of the next generation machine, you say that at the respective company not only development, it considers also the choices which use the tip/chip of the GPU vendor of PC type.

Thus, Nvidia technologies must be so horrible that Sony decides to pen contracts, pay out royalties and link their future PS line with Nvidia, all for some lowly leashed road sweeper picking up the leftovers!


Also Watch thinks Nvidia will provide audio video and network for PS3 like MCPX.

"Media processor business is very important even in home appliance field. We, with the SONY computer entertainment and the other two partners, make PlayStation of the next generation ", (Huang)

Huang, GPU and MCP (Media & Communications Processor) and (Wireless) classifying product category of the same company, into 3 of Media Processor. The cooperation with SCEI was explained other than GoForce as 3rd Media Processor. As for GPU specialization, as for MCP specializing in the media & communication & security, it is combined with GPU in graphics. Vis-a-vis that Media Processor, when you look at the example of GoForce, takes charge of the media processing of all round with the programmable hardware not only graphics. In addition, the Hisashi of SCEI 夛 is good from the time before the wooden person the companion tip/chip which is combined with the Cell processor which is CPU of PS3 was not the graphic engine and it called the "media engine". From these speeches, the tip/chip of NVIDIA is presumed that it is the GPU + MCP, media processor. In other words, it is seen that it mounts also the processing function of non graphics such as audio and network.

Remember deadmeat media processor topic (http://www.beyond3d.com/forum/viewtopic.php?t=10300&highlight=media+processor)?

[one]

Sony fans here have breath a sigh of relief and excited with Nvidia audio and video solution.

Also Watch thinks Nvidia will provide audio video and network for PS3 like MCPX.

That page only says that the end product which nVIDIA co-develops will be a media processor type product such as GoForce, not a separated GPU/IGP nor a separated MCP. In other words, it's not clear if nVIDIA will provide "audio video and network" for PS3. IMO it's natural that nVIDIA offered only 3D-related IP for the media processorm, for they mentioned specifically "GeForce technology".


Remember deadmeat media processor topic (http://www.beyond3d.com/forum/viewtopic.php?t=10300&highlight=media+processor)?

Well, have you read the answer by Panajev2001a to the topic starter? Also, it looks like this Deadmeat thought "various media processors" as things that compete against Cell in the Sony inner politics. But the reality is, Cell and those "various media processors" are synergistic in Sony products.

[rabidrabbit]

Who is that "Wooden person" that often appears in these texts about Sony and PS? Is Kutaragi just a "puppet" and this "wooden person" is really the brains behind all PS... or is Kutaragi this "wooden person" like a puppet is a "wooden person"???

And that DM's thread... what has it got to do with this thread?
I really couldn't be bothered to read the whole thread, but what he seems to argue there is that Cell wouldn't be in PS3 in any form (although in some of his postings he agrees that "Cell" will be there, but with separate "Media processor", like it will be, and which was already agreed by the so called "Sony fans" in that thread :? )

And... it's not the wisest idea to bring another flame-infested thread to a topic that is in itself ruined already, unless you just need some backing up by your (believed to be pased, but apparently resurrected) idol :(

[pahcman]

Thats what Watch thinks not me. So far i not bothering to figure out all possible combinations, all i know is, Sony chose Nvidia.

The other topic link is not about deadmeat saying Cell will compete with Media Processors, but with where Watch sees with Sony PR Nvidia PR and Media Processors(IIRC mp is a buzz word invented by Nvidia too).

[rabidrabbit]

Did they already get the "Chap to English" translating option ready in Babelfish :?
Edit: A horrible possibility just occured in my mind! Could chap and Deadmeat be using the same account for posting, as at times pachmans postings are understanable english, and at times they are gibberish :shock:

[one]

Thats what Watch thinks not me. So far i not bothering to figure out all possible combinations, all i know is, Sony chose Nvidia.

The other topic link is not about deadmeat saying Cell will compete with Media Processors, but with where Watch sees with Sony PR Nvidia PR and Media Processors(IIRC mp is a buzz word invented by Nvidia too).

Okay, okay :lol:
When I joined this board Deadmeat was already banned IIRC as I saw the sticky topic about him, but I assume it was funny to read his comments realtime :P
Select quotes from the thread:
You expect CELL to substitute A/D and D/A converters too ?

They already said that they would still use dedicated Silicon in CELL based systems for things that were not worth doing in Software on CELL ( example would be Texture Filtering and Sampling ).

Do not increase the scope of CELL beyond its intended one.
Something needs to have a CRTC inside the PS3 to output to Monitor.. Something Broadband Engine will not have.

These "media processors" are probably GPU's..

[london-boy]

Okay, okay :lol:
When I joined this board Deadmeat was already banned IIRC as I saw the sticky topic about him, but I assume it was funny to read his comments realtime :P
Select quotes from the thread:


Oh it gets funnier, with the resurrected DM...

Don't worry about it we ain't goin anywhere

Sammy didn't buy us to toss over to someone else.

Are we gonna use the series 5 or a custom PVR chip in our next board?

That's for me to know and for you to found out.

I've only told a couple of my closest friends, friends whom I can trust.

I will tell you guys something and that is, we ain't being bought by noone just use alittle bit of brain people, how could we? Sammy is our new Parent company now.

It's as if people just missed the whole saga that took place within the last 6 months.

That's all you get out of me.


AND THE BEST EVAH exhange of posts (http://www.beyond3d.com/forum/viewtopic.php?p=277547#277547)


Sorry, very off topic.

[darkblu]

not to say that superscalar processors are all that great at getting more than 1 instruction executed per clock thanks to memory latency and other hazards. I think someone gave an average of .8 instructions per clock for a K8. If the old chips weren't orders of magnitude worse, modern chips would look much less impressive.

well, for one the 386 was surely not the apogee of the single-scalar cpus :)

[3dilettante]

not to say that superscalar processors are all that great at getting more than 1 instruction executed per clock thanks to memory latency and other hazards. I think someone gave an average of .8 instructions per clock for a K8. If the old chips weren't orders of magnitude worse, modern chips would look much less impressive.

well, for one the 386 was surely not the apogee of the single-scalar cpus :)

Definitely not, though it was in the good old days where processors didn't outclock their RAM by over an order of magnitude.

Those were the days.

[Fox5]

Didn't nV do the Saturn video and audio, as well?

Apparently no.

http://www.nvnews.net/vbulletin/printthread.php?t=29935

no no no, Nvidia never made any chips for the Saturn at all. nor where they going to. the Sega-Nvidia partnership was for a chip for a successor to the Saturn. whatever that sucessor was, it went unmentioned, and is also not the Dreamcast.

the Saturn was in development in 1991-1992, before Nvidia even existed. Saturn uses Hitachi CPUs and SEGA-designed 2D sprite & background graphics processors (VDP1 and VDP2) which are derived from Sega's System32 arcade board of 1991. plus Saturn has a bunch of other processors, from Sega, Yamahi.

there is no Nvidia technology in Saturn whatsoever. Nvidia was a new startup company by the time the Saturn was finished (1994).

Nvidia's first chip, the NV1 of 1995, was used in the Diamond Edge 3D card. Sega and Nvidia had another agreement where Sega Saturn games were ported over to the NV1 chip / Diamond Edge 3D card, and Sega Saturn controllers could be plugged into the card--the Diamond Edge 3D card was a multi media card with 3D, 2D, Audio, etc.

the NV2 chip that Nvidia was working on for Sega used quads, like the NV1, but Sega didn't want quads. Nvidia refused to make a chip that pushed triangles. Sega backed away from Nvidia without any drama at all (unlike the well known Sega-3Dfx dispute)

the NV2 chip was scrapped, although it very well might have been used in the Sega PICO. but there's no evidence of that, though (NV2 in PICO comes from the FiringSquad.com article)

After Sega parted ways with Nvidia sometime in 1995, Sega went on to talks with Lockheed Martin over using Real3D in a console. nothing came of those talks. then Sega went to PowerVR and 3Dfx, ultimately selecting the PowerVR2 based Dural in 1997 which was then named Katana, and then finally, Dreamcast 1998.


What!? Everything I'd heard was that the sega saturn was basically a higher clocked 32x, except the nvidia nv1 was thrown in shortly before release, used quads(just as the nv2 was planned to, and the saturn used quads), and that the nv1 based cards on pc were the only ones with hardware accelation for saturn ports.
So if saturn was only a higher clocked 32x with a cd rom drive how come its games look so much better? Or maybe they don't, first gen saturn games didn't look very good, but 32x 3d games generally looked like stuff that could be done with a superfx chip on the snes.

BTW, wasn't the PICO only a 2d system?

[london-boy]

So if saturn was only a higher clocked 32x with a cd rom drive how come its games look so much better? Or maybe they don't, first gen saturn games didn't look very good, but 32x 3d games generally looked like stuff that could be done with a superfx chip on the snes.


:shock: Errr, i would NEVER have thought i'd defend the 32X in my whole life, but i'm sorry that is just plain incorrect, by a large margin.

Maybe you should check.

[Fox5]

So if saturn was only a higher clocked 32x with a cd rom drive how come its games look so much better? Or maybe they don't, first gen saturn games didn't look very good, but 32x 3d games generally looked like stuff that could be done with a superfx chip on the snes.


:shock: Errr, i would NEVER have thought i'd defend the 32X in my whole life, but i'm sorry that is just plain incorrect, by a large margin.

Maybe you should check.

I own a 32x, and I'll defend it's 2d capabilities but 3d seems a bit lacking...(though it seems strange for me to take that position, since you can always say the 32x holds at least a small advantage over the snes's 3d, yet not always the 2d...just the 2d can look much more impressive than the 3d)
http://www.defunctgames.com/pic/knuckleschaotix32x2.jpg

Ok, better than starfox on snes, but how about fx2 chip games like the unreleased starfox 2?

http://www.dcshooters.co.uk/sega/32x/images/vf.png

http://www.emulationgalaxy.co.yu/images/VR-32X.jpg
I believe this is the 32x version of virtua racing...

http://www.emulationgalaxy.co.yu/images/systems/MD2_Virtua_Racing_Deluxe.jpg

http://www.althena.com/images/genesis/virtuaracing2.jpg
And this is the genesis.

http://www.dcshooters.co.uk/sega/32x/images/starwars.png
This is more like starfox quality though.(I was going to use shadow squadron, but it looks much better)

[GwymWeepa]

lol I owned a 32x as well, man what a rip off, though I did have fun renting Virtua Fighter and Virtua racing.

[jvd]

The 32x should never have been released although i do love my virtual racing and star wars on it.


If the 32x was never released the world might still have had sega as a hardware manufacturer

[see colon]

the 32x had dual sh2's just like the saturn did, and they were indeed clock slower than the saturns, but it lacked several components that the saturn had (vdp1 and vdp2 for example) that gave the saturn a huge advantage.

the 32x's 3d speed was much better tha either fx or fx2 equiped snes games. check out doom, for example. the snes version had no textures on ceilings or floors, and still ran at half the frame rate. star fox is a choppy mess on the snes. star wars is pretty smooth on the 32x.