If PS3 can really do 1Tflops

[clem64]

won't it completely destroy GC2 and Xbox2 in the graphics department?

I mean, PS2 can do 6.2 Gflops. GT3 creator said that the game is using 25% of the CPU's power. Now GC and Xbox can do a similar amount of Gflops. I'm guessing about 1.5 - 2.0 in-game (somebody correct me if I'm wrong).

How many gflops will the CPUs in GC2 and Xbox2 be able to do? 15-20 in-game I would guess? Compared to, say, 25% of 1Tflops in a real gaming environment. Will their GPUs be able to make up for a 200+ Gflops disadvantage?

 

[randycat99]

MIND YOU ALL- He said if it can do it. So let's not spend 6 or 7 pages on how it can or cannot be realized in hardware. This is about speculation here based on a given baseline.

 

[V3]

won't it completely destroy GC2 and Xbox2 in the graphics department?

No it won't.

 

[hughJ]

if the ps3 delivers on it's 1TFLOP spec.

and

if the competing systems use off-the-shelf/consumer-level CPUs.

and

IF the graphics chipsets on those systems offer ZERO innovation in how well they perform; how well their architectures are geared in doing what would otherwise have to be CPU-related functions.

then yes, am sure the ps3 would completely destroy its competitors.. but assuming that the role between CPU and 'GPU' will remain 100% static is rather foolish. The path Sony is taking is completely different than that of the Nintendo and Microsoft, simply comparing CPU power really doesn't tell the whole story.

 

[randycat99]

On the contrary, GPU's are moving closer and closer to flexible programmability and flexibility. Sony's approach represents an ultimate expression of that- in effect, the GPU is a CPU, yet radically parallelized for GPU-like performance. It's possible that with whatever Xbox2 and GC2 come up with in GPU silicon, PS3 will be able to duplicate in operation with its extremely configurable computation units and sheer brute force of its promised throughput. Well, that's the strategy, at least. Will it work out? Only time can tell...

 

[hughJ]

But there's also a happy median in there somewhere, where the GPUs are still able to do many things far more efficiently than the duplicated result via software. Regardless of the programmability of the GPU, they're still quite centralized on the graphics portion of the overall picture, where as the EE2 (or whatever) will be required to be much more general-purpose.

 

[randycat99]

Relative efficiency won't be relevant unless you happen to be running short on resources, of which the PS3 is intended to have plenty to burn. Equally, there will be certain combinations/types of effects that could end up being very difficult to implement in a "frozen" GPU design that could be simplified, integrated, and streamlined to run quite effortlessly on the PS3. Not that this distinction will really be important in the next generation as all 3 offerings will probably be so powerful as to make just about anything you want possible (but you never know what graphics people will think of trying). It's even more moot when you consider that the Cell architecture will be fast for the very same reasons that a GPU is fast, from an architectural standpoint.

 

[speng]

One thing to keep in mind.

There is power and energy everywhere in this world (ocean, volcano, solar energy) but it cannot be used efficiently. Why? because there are no tools to properly take advantage of these powerful phenomenons.

Even with 1TFLOP of processing power, it's the tools, APIs and development kits that determine how well the programs that are developed for use will run.

If Sony, IBM and Toshiba can deliver the power and easy to use powerful tools, then both Nintendo and Microsoft have some stiff competition coming.

Big if, which I have very little doubt they will pull it off. IBM is a beast when it comes to R&D and software development.

Sony cannot afford to lose on the Cell technology. It is not only for use in this game, it's for use in multiple electronic equipment. IBM is also banking on these things as is Toshiba.

I forsee a powerful chip, I'm only concerned about the quality of the tools that come with it.

Speng.

 

[hughJ]

Relative efficiency won't be relevant unless you happen to be running short on resources...

...all 3 offerings will probably be so powerful as to make just about anything you want possible

There will [essentially]always be limitations that are quite quickly and easily reached. Relative efficiency most certainly will be relevant as a result.

Equally, there will be certain combinations/types of effects that could end up being very difficult to implement in a "frozen" GPU design that could be simplified, integrated, and streamlined to run quite effortlessly on the PS3.

Yup, it's very likely that the GPU would be less versatile, with some effects that would be more-suited("simplified", "integrated", "streamlined", "effortlessly") to the PS3, the question then would be whether or not that advantage would be enough to overcome the likely disadvantages of inefficiency when compared to more-native 3D architecture.

To get back to the topic of the thread; I doubt either design choice will offer any clearly superior results over their competitors. So likewise, I highly doubt the PS3 will/would, "completely destroy GC2 and Xbox2 in the graphics department."

 

[London Geezer]

Relative efficiency won't be relevant unless you happen to be running short on resources...

...all 3 offerings will probably be so powerful as to make just about anything you want possible

There will [essentially]always be limitations that are quite quickly and easily reached. Relative efficiency most certainly will be relevant as a result.

Equally, there will be certain combinations/types of effects that could end up being very difficult to implement in a "frozen" GPU design that could be simplified, integrated, and streamlined to run quite effortlessly on the PS3.

Yup, it's very likely that the GPU would be less versatile, with some effects that would be more-suited("simplified", "integrated", "streamlined", "effortlessly") to the PS3, the question then would be whether or not that advantage would be enough to overcome the likely disadvantages of inefficiency when compared to more-native 3D architecture.

To get back to the topic of the thread; I doubt either design choice will offer any clearly superior results over their competitors. So likewise, I highly doubt the PS3 will/would, "completely destroy GC2 and Xbox2 in the graphics department."

oh, also, u should remember that IF it CAN do 1Tflop then it wont be 1000x PS2. it would *only* be 156.25X the power of PS2....
so what are we going to believe,
1)a ps3 thats 1000x PS2

or

2)a PS3 that can do 1Tflop????


the second option would be quite an *underpowered* platform compared to the first option. to be precise it would be about 6 times slower..... and thats quite a lot..............

just giving u some material to speculate on

 

[McFly]

I do not remember which number came from Sony (If any), but if I'm not wrong then it was 1000 times faster. So it will be 4 to 6 TFlops in my opinion.

Of course just speculation.

Fredi

 

[Phil]

oh, also, u should remember that IF it CAN do 1Tflop then it wont be 1000x PS2. it would *only* be 156.25X the power of PS2....
so what are we going to believe,
1)a ps3 thats 1000x PS2

or

2)a PS3 that can do 1Tflop????


the second option would be quite an *underpowered* platform compared to the first option. to be precise it would be about 6 times slower..... and thats quite a lot..............

just giving u some material to speculate on

Well, AFAIR, both were mentioned. One thing that's not sure though, was how it was ment. GFLOPs isn't the end-performance of a console, but what you get on screen etc. Perhaps getting 1 TFLOP out of PS3 will be short of being 1000 times more, but it certainly would be a remarkable achievement given the time and the R&D that went into achieving such a thing. Look at it like this, GFLOPs isn't everything just aswell polygons aren't everything. Upon looking at the console as a whole, I'm sure PS3 will deliever a 1000 times improvement.

 

[London Geezer]

oh, also, u should remember that IF it CAN do 1Tflop then it wont be 1000x PS2. it would *only* be 156.25X the power of PS2....
so what are we going to believe,
1)a ps3 thats 1000x PS2

or

2)a PS3 that can do 1Tflop????


the second option would be quite an *underpowered* platform compared to the first option. to be precise it would be about 6 times slower..... and thats quite a lot..............

just giving u some material to speculate on

Well, AFAIR, both were mentioned. One thing that's not sure though, was how it was ment. GFLOPs isn't the end-performance of a console, but what you get on screen etc. Perhaps getting 1 TFLOP out of PS3 will be short of being 1000 times more, but it certainly would be a remarkable achievement given the time and the R&D that went into achieving such a thing. Look at it like this, GFLOPs isn't everything just aswell polygons aren't everything. Upon looking at the console as a whole, I'm sure PS3 will deliever a 1000 times improvement.

mmmmm i dont know........ i like speculating though

i just want to SEE what kind of images such a beast would produce.......

can u imagine if they keep a PS2-like interlaced shitty output for ps3!!!! that would be HILARIOUS!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

er....

 

[BenSkywalker]

On the contrary, GPU's are moving closer and closer to flexible programmability and flexibility. Sony's approach represents an ultimate expression of that- in effect, the GPU is a CPU, yet radically parallelized for GPU-like performance.

1TFLOP by '05 could be integrated graphics level(cheap ass eMachines caliber) for GPUs though, they aren't far off from that number now.

If the PS3 has a 1TFLOP chip and a rasterizer equal to that of the GS then it would get its ass kicked in the visual department(I wouldn't be surprised if the R350/NV30 w/P4 3GHZ/Barton 3000 could numble it), although obviously they will be improving that significantly too. If they have a solid DX9+ level rasterizer paired a TFLOP CPU they might be able to pull even in theoretical graphics output. DX10 level with TFLOP should have them quite competitive if not pulling ahead.

That is speaking in a theoretical sense of course, artists are going to be the biggest limiting factor in next gen console graphics.

 

[BoddoZerg]

Personally, I don't have one shred of faith in Sony's Cell hype. It sounds like technological mumbo-jumbo mixed with shades of Transmeta and a dash of Ginger. (Remember... handheld Crusoes were going to be more powerful than any Pentium!) Namely, they are releasing huge amounts of speculative big-numbers (1000x! TeraFLOP! More power than a SuperComputer!) waaaaaay ahead of time.

It's simply bad practice to build up so much advance-hype for a technological advance, because they are so unpredictable by nature. Just look at what happened to the NV30. Even NV30 was only a little more than a year of pre-hype. With PS3 we are talking about two and a half to three years of pre-hype. That's just not a viable recipe for accurate information.

 

[London Geezer]

Personally, I don't have one shred of faith in Sony's Cell hype. It sounds like technological mumbo-jumbo mixed with shades of Transmeta and a dash of Ginger. (Remember... handheld Crusoes were going to be more powerful than any Pentium!) Namely, they are releasing huge amounts of speculative big-numbers (1000x! TeraFLOP! More power than a SuperComputer!) waaaaaay ahead of time.

It's simply bad practice to build up so much advance-hype for a technological advance, because they are so unpredictable by nature. Just look at what happened to the NV30. Even NV30 was only a little more than a year of pre-hype. With PS3 we are talking about two and a half to three years of pre-hype. That's just not a viable recipe for accurate information.

The thing is, the console market and the High-end video cards market are TOTALLY different.....

remember that someone buying an NV30 KNOWS why he's buying it and why choosing either that or an R300.

in the console business it does not work like that. and i think we've seen that already with PS1 and PS2....

 

[Panajev2001a]

1TFLOP by '05 could be integrated graphics level(cheap ass eMachines caliber) for GPUs though, they aren't far off from that number now.

of course... the famous NVFLOPS...

Fact is I do not believe the Rendering core will be as simple as you perceive it... fact is the GPU will be programmable ( you have seen the patent and you know we have 64 Integer Units and 64 FP Units in the Visualizer in addition to the 4 Pixel Engines and the Image Cache )...

Fact is, you are basically saying "well if they had a basically purely software Rasterizer and T&L engine 1 TFLOPS would be e-machines level"...

Of course if you took the current GS even and counted the number of flops for the set-up engine and you rated in a MIPS scale the power of all the rest of the GS's rendering pipeline ( integer pipeline ) you would obtain a very high number... GeForce FX is already over 250 GFLOPS isn't it ?

So, in 1999 they had ready a machine that kicked the living crap of high end PCs in RAW performance and npw in 2005 they ship a console that is as powerfule as a 2005 e-machine ?

 

[Vince]

1TFLOP by '05 could be integrated graphics level(cheap ass eMachines caliber) for GPUs though, they aren't far off from that number now.

I'd say there's a distinct diffrence between talking about the MPU as outlined in that patent and comparing it to a GPU - which even at the DX9 level is pretty static and lags behind even the PS2 at certain, albiet few, functions which require a high degree of architectural programmability.

Going by that patent; I'd say comparing the TFlop Broadband Engine, or whatever name marketing throws on it, would be better compared to competing platorms when viewed either as: (a) Just the MPUs (b) Part of a complete system. Not comparing a MPU/CPU like device against a rasterizer.

If the PS3 has a 1TFLOP chip and a rasterizer equal to that of the GS then it would get its ass kicked in the visual department

Ok, now this I have to question. I've stated it many times here before; and these ideas were just stated by ATi's Orton, so I must ask yet again. Why would persuing a micro-polygon route be inferior than the more PC-centric system which faces namely storage and bandwith constraints?

In any case, the nororious patent does contain significant computational resources, not unlike those found on the MPU/BE, in the 'front-end' of the visualizer. Thus, I'd tend to feel that the computational capabilities are significant and programmable enough to compete in a relative shader implimentation and if it has a respectable sampling rate in addition- would it be out of line to say its the closer praxis of the PRman ideology, as opposed to the PC's current design?

 

[Vince]

GeForce FX is already over 250 GFLOPS isn't it ?

I believe the Nv30 has roughly 250GFlops aggrigate; with around 50 of that in the fragment shader IIRC. If this means anyting tangible.

 

[BoddoZerg]

The thing is, the console market and the High-end video cards market are TOTALLY different.....

remember that someone buying an NV30 KNOWS why he's buying it and why choosing either that or an R300.

in the console business it does not work like that. and i think we've seen that already with PS1 and PS2....

In any case, I just feel that the sheer scale of pre-hype we've been seeing for PS3 is completely unwarranted and can only lead to trouble.

You don't see MS and Nintendo hawking Teraflops and Gigapixels for the Xbox2 and GC2, do you? The common answer is - "Well they haven't settled on a final design". My reply is this - The XBX2 and NGC2 are set on the same timescale as the PS3. If the first two don't have a completely finalized design, what makes you believe the third does? Having a generalized outline for multiprocessor cells is very different from having everything set in silicon with defined functionality and known clockspeeds.

The thing is, it's not just nVidia and other 3d graphics people. You also don't see Intel and AMD quoting Teraflops for a processor to be released in 2005. You'll see them talking about advances like 64-bit CPUs and 60 nm processes... but you don't see them quoting a 1000x increase in CPU power, or using phrases like "as powerful as all the computers in the world combined". That kind of performance-quoting pre-hype is generally reserved for the likes of Bitboys, pre-product launch Transmeta, or Deam Kamen's mysterious "Ginger". It is simply not a respectable business practice and I despise it.