PlayStation III Architecture

How do you utilsed the performance than ?


How ? Well, by doing what I wrote in the rest of the post...

It is true when we write code ( write != compile ) we do not think about the processing speed of the machine and what we worry about is synchronization of the different APUs and different parts of the task we are executing, etc...

We basically synchronize on the absolute timer... "timer fires => APU is woken up => result is expected" sounds like multi-tasking on a modern OS, but with some of its mechanism ( of the process/task/thread scheduler ) built in HW or so it seems...

As I wrote at the beginning of this message I do believe that in the code we have a way of specifying the speed of the CPU: how else could we talk about code written "for an older/slower APU" running on a "newer/faster APU"... this also implies we can write code that gets compiled in such a way we can define it as being "tailored for a newer/faster APU"...

What changes should the amount of operations you can do for each task...

If you know you are writing code for a 8 GHz Cell device and we assume the time slice given by the absolute timer has been remaining the same since the first Cell ( "the time slice given by the absolute timer will not change" ) than we will know ( and the compiler will too ) that we can fit more work in each time slice offered by the absolute timer so that we can reach a higher level of utilization for our HW ( we are not interested about power consumption, in this example, but about performance ).


You rightly wondered what I meant by a simple Cell chip for Cell based TVs...

I read an article on EETimes magazine yonks ago, it was a Toshiba engineering talking about TV of the future. One of the point I remember was that in Sports like Football, they wanted to let the viewer to be able to replay a goal, from different angels, and one way to do this is to recreate the scene in 3D, and let the viewer manipulate it in real time. He estimates that it requires something around 8 billion polygon/s to get something somewhat realistic.

So I don't know why you think TV would not need such a fast chip is beyond me.

Remember high end TV has alot of margins, they can include state of the art chip in there.

I am sorry, my analisys was confined to medium-low to medium-high priced TVs and not $12k HDTVs so I was not going to think about multi-billions polygons per second...

8 billions polygons/s... uhm... a bit high even for the mighty PS3... say 200-400 MPolygons/s and I'd say maybe ;)

Still, maybe we will see more powerful than PS3 chipsets powering very high-end TVs in the future.. maybe they will play PS3 games as well... who knows...







You asked about Linux... well I think Linux is good enough... more than good enough to serve as the basis for Cell OS...

Sony is quite fond of Linux, but IBM is even more and their multi-billion dollars investment in Linux related R&D ( you can do a simple search at IBM's website and you will find quite a good amount of links ) and you will see why Linux could really have a future in Sony, IBM and Toshibas plans :)
 
If you know you are writing code for a 8 GHz Cell device and we assume the time slice given by the absolute timer has been remaining the same since the first Cell ( "the time slice given by the absolute timer will not change" ) than we will know ( and the compiler will too ) that we can fit more work in each time slice offered by the absolute timer so that we can reach a higher level of utilization for our HW ( we are not interested about power consumption, in this example, but about performance ).

Aww sorry, I totally misread this one :oops: my fault.

See my problem with this, is that absolute timer will dictate the sequential synchronisation of your data. If you don't change the absolute timer your parallelsation will improve, but your sequential speed will remain the same.

So you'll hit that Amhdal's Law.

I am sorry, my analisys was confined to medium-low to medium-high priced TVs and not $12k HDTVs so I was not going to think about multi-billions polygons per second...

8 billions polygons/s... uhm... a bit high even for the mighty PS3... say 200-400 MPolygons/s and I'd say maybe

Yeah he was talking about 2011. Anyway even in low end TV, those chip need to at least function as digital recorder of somesort. So you don't need setop boxes anymore.

You asked about Linux... well I think Linux is good enough... more than good enough to serve as the basis for Cell OS...

Sony is quite fond of Linux, but IBM is even more and their multi-billion dollars investment in Linux related R&D ( you can do a simple search at IBM's website and you will find quite a good amount of links ) and you will see why Linux could really have a future in Sony, IBM and Toshibas plans

Hehehe, that would be great :)
 
Aww sorry, I totally misread this one :oops: my fault.

No problem V3, it is ok :D

See my problem with this, is that absolute timer will dictate the sequential synchronisation of your data. If you don't change the absolute timer your parallelsation will improve, but your sequential speed will remain the same.

So you'll hit that Amhdal's Law

Who knows... who knows when they will hit it... There are areas in which parallel computation has a lot to go ( 3D graphics, physics, A.I., server request processing, etc... ) even after PS3 and besides we are reaching the point in which we cannot really push sequential thread speed THAT much faster ( Intel speed-demon approach ).

I understand your point and I think that since the day this patent was written till the day the finalised specs for Cell will be completed 100% this has been one of those areas which have been reviewed and reworked till a very good solution was found... I expect changes in areas like that...
 
Just another piece of news, from a reliable source.


One month from now, Sony Computer Entertainment will reveal the technical specifications of its PlayStation 3 console. This year?s Games Developers Conference, to be held in San Jose, USA, between March 4-8, is the time and place for one of the most exciting announcements this side of Nintendo?s Game Boy Advance SP being endowed with a screen that is actually useful for playing games!
The announcement is expected to be made by SCE manager of developer relations Mark DeLoura, though this could not be confirmed.
The news was leaked to us from the highest-level third-party development sources available. The credibility and standing of the individual who told us this convinces us that it is correct.
What?s more, this revelation has sparked intense rumour that Sony will make a consumer announcement at this year?s E3. It is thought that a full consumer hardware plan will be unveiled, with a Japanese launch coming within 18 months of the show.


http://spong.com/x?art=4401

Could it be true? Or is it just the unstoppable Sony Hype Machine Working again????
 
If it's true and if it will come out late 2004 and will have more then 1TFlops and third parties already working on games, then ... then MS and big N would have to figh one year later against a huge userbase similar to this generation and loose again.

But I don't think it's true. Mid 2005 or late 05 is early enough for the next generation if you ask me.

Fredi
 
The schedule is very similar to the ps2 lauch from a japanese point of view. The ps2 was supposed to launch Christmas 99, after having being announced in March 98, and was delayed 3 months.
 
PC-Engine said:
With that many ifs, time travel would be possible too

Actually, aren't you doing that right now? <ducks> ;)


You'd think at E3 for sure, or even GDC '03 unveiling, that it would be presented by the Trifecta of Kutaragi, Okamoto and Kaz.

Bring Kaz along for the fun. I'm convinced they cut off his caffiene for a few weeks prior, then give him some espresso (or Redbull & Jager) before unleashing him on E3 to claim Playstation is the winner before it's launched or some other crazed taunts.
 
Vince... you want to see Kaz unleashed ?

Watch out for what you ask and you will see MS would conter-attack unleashing their own leech, Steve "Monkey Dance" Ballmer!!!!!!!
 
The schedule is very similar to the ps2 lauch from a japanese point of view. The ps2 was supposed to launch Christmas 99, after having being announced in March 98, and was delayed 3 months.

U sure, I think the announcement was march 1999, the day the playstation conference, to celebrate the sale of the 50th million psone, took place...
 
Main Processor "Cell" on .10 micron process produced by IBM.
256 Bit
Clock speed 4Ghz
Floating point performance: 1 Trillion Floating point operations per second.
16 RISC cores on single DIE


Main system ram 512MB "yellowstone" memory

Graphics

Graphics Synthesizer 3

Clock speed 825Mhz
Vram 128mb (Yellowstone?)
Polygon performance: 6 billion polygons per second (raw)
With effects: 2-3 Billion polygons per second
Particle performance 5 billion particles per second
16X anti aliasing (Hardware)
26 Texture pass's per clock cycle.

Supports: Mip mapping, FSAA, polygon distortion, volumetrics, lighting, multi pass rendering, tri-linear rendering, bump mapping.


Storage medium

Sony developed "Blu Ray" disks compatable with Standard Dvd rom disks.
60GB HDD using Fiber optic transfer
Built in 10/100 network adapter
4 controller ports using USB 2.0 specifications


Sound

'Spruz 2' sound system.

512 simutaneous 2d voices
128 simutaneous 3d voices
Fully compatable with Dolby and beyond.

Notes

Graphics that of a Low budget CG movie. Online gaming right out of the box, microphone support, broadband only. System draws power from Cores in the Cell, on a task related basis.
 
As said many a time, PS3 specs won't be given until 2004 at the soonest. in all likeihood. probably sometime between March and Dec 2004.

launch will be March 2005 in Japan at the soonest. anywhere from March 2005 to March 2006. with U.S. release in fall 2005 to fall 2006.
 
marconelly! said:
Just noticed this:

Just another piece of news, from a reliable source.
.
.
.

http://[b]spong.com[/b]/x?art=4401

:LOL:

You must be very new to this whole internet gaming news thing. Spong is basically a National Enquirer of gaming news.

Even if I was so new, with just some months I would have had enough Megatons¡¡¡:D:D

I should have written "reliable" to make clearer my intention, which I thought was clear enough when I wrote:
"Could it be true? Or is it just the unstoppable Sony Hype Machine Working again????"
 
megateto said:
marconelly! said:
Just noticed this:

Just another piece of news, from a reliable source.
.
.
.

http://[b]spong.com[/b]/x?art=4401

:LOL:

You must be very new to this whole internet gaming news thing. Spong is basically a National Enquirer of gaming news.

Even if I was so new, with just some months I would have had enough Megatons¡¡¡:D:D

I should have written "reliable" to make clearer my intention, which I thought was clear enough when I wrote:
"Could it be true? Or is it just the unstoppable Sony Hype Machine Working again????"



i really don't know why everyone's always nagging about this *sony hype machine* blah blah blah...... i mean at the end of the day when launching the PS2 all they did was unveiling the Specs. they said PS2 could do 66 million polygons RAW (and they specified it) and at the end of the day that is real performance. of course u wouldnt have a game comprised of only flat shaded polygons but they did specify RAW poly performance... they also came up with the 6.4GFlops TOP performance. everyone knew that these were only SPECS.... its the fanboys who hyped the machine as the next supercomputer or things like that (remember the ridiculous rumours about china and irak using ps2's to guide missiles.....)...

Sony did not hype its machine any more than Microsoft or Nintendo did..... And Remember that Microsoft also came up with the 5x more powerful (whoich later became 3x more powerful) than anything the universa had ever created of crap like that... and they also said that by 2005 there would be like 100 million Xbox's sold on the planet.... :LOL:

really, this whole *hype machine* crap is just ridiculous. Everyone does it and to be honest many people do it more than sony.
 
london-boy said:
i really don't know why everyone's always nagging about this *sony hype machine* blah blah blah...... i mean at the end of the day when launching the PS2 all they did was unveiling the Specs. they said PS2 could do 66 million polygons RAW (and they specified it) and at the end of the day that is real performance. of course u wouldnt have a game comprised of only flat shaded polygons but they did specify RAW poly performance... they also came up with the 6.4GFlops TOP performance. everyone knew that these were only SPECS.... its the fanboys who hyped the machine as the next supercomputer or things like that (remember the ridiculous rumours about china and irak using ps2's to guide missiles.....)...

Sony did not hype its machine any more than Microsoft or Nintendo did..... And Remember that Microsoft also came up with the 5x more powerful (whoich later became 3x more powerful) than anything the universa had ever created of crap like that... and they also said that by 2005 there would be like 100 million Xbox's sold on the planet.... :LOL:

really, this whole *hype machine* crap is just ridiculous. Everyone does it and to be honest many people do it more than sony.

Hey, if I rember well, Nintendo was the only one to specify a conservative estimate of real game environment causing people in some fora to bash Nintendo. :LOL:
 
Take it easy, friend.. I was only joking. In fact, I was thinking about that very same thing some weeks ago when I was reading an old Edge issue (the one in which they covered the unveilment of the PS2), and although they mention the 66 figure, they also say something about real game conditions, and talk about 16 or less.. I'd better re-read it (for the third or fourth time¡¡).
 
Back
Top