CONFIRMED: PS3 to use "Nvidia-based Graphics processor&

You do realise that if your only concern is developing hi-res art assets etc. etc. a PS3 developer can easily create all their art content using XNA on windows and then transfer the assets to a PS3 dev kit and use the results in their own PS3 engines?

 

Nothing a contract and a watermark couldn't prevent.

 

True, Microsoft might like that, but would middleware developers agree? A lot of companies making modelling an animation tools, for example, don't make them for each console. They make them for PC, and developers use them on PC to create their art assets. I doubt such tools companies would want to engage in XNA if it limited the use of their product's output to windows.

Of course, I'm not saying everything about XNA will be useful for people working with platforms other than windows. Obviously any code that calls XNA-related functions etc. to bind two engines together, for example, isn't going to work outside of windows. But on the content/art creation side, there's potential for anyone to take advantage, regardless of what platform that content/art ends up on.

Also, how do you watermark a 3D mesh (for example)? (Honest question, i'm not suggesting it's impossible). And how does the program know you are going to take it and use it on another platform? It can't watermark all output.

 

rambus memory is too slow for gpu(random access), must have big embended mem(32MB?)
but this emb memory size is too small for texture and geometry
hence dont vertex processing in gpu, only cpu

 

I don't think it will be that hard to directly expose that capability in the APU ISA.
There's much more than that if you read the patent. APUs can make DMA requests and group them..checking later for group(s) of dma requests completition.

SPUs have a L1 cache! and we really don't know how slow external memory references will be.
The real problem (me and Faf are going to repeat that 1000 times..) is what APUs can do while waiting DMA queries to be completed.

ciao,
Marco

 

Funny, I do too. Are they in the midwest by chance?

 

Yes and no... I believe this is why they want their console to be compatible with Linux, instead. Therefore, leaving Linux to the job for them. One things for certain, though, this is where Microsoft are at their strongest.

 

Here's something to take a look at
http://www.eet.com/semi/news/showAr...BCCKH0CJUMEKJVN?articleId=54200580&pgno=2

but more specifically.....
But UNC's Zimmons has his doubts. "I believe that while theoretically having a large number of transistors enables teraflops-class performance, the PS3 [Playstation 3] will not be able to deliver this kind of power to the consumer," he wrote in response to an e-mail query from EE Times. "The PS3 memory is rumored to be able to transfer around 100 Gbytes/second, which would mean it could process new data at roughly 25 Gflops (at 32 bits) — far from the 1-Tflops number."

 

... would be the best way to make my first comment in Console Talk. I surely didn't see this coming!

 

Please, not again. :?

Fredi

 

Which is probably why Sony has taken part in helping Nvidia with the GPU.

 

Just to address the earlier comment of MS using their $40+ billion in cash to buy Sony. It's not really feasible due to anti-trust laws, and it would be more efficient to simply buy Rockstar, Konami, Capcom, and Square-Enix for around $8 billion. Taking away GTA, MGS, RE + DMC, and FF/DQ would be the easiest way to hurt PS3. It's still a difficult proposition though.

 

Guden Oden:
To me your focusing on the problems at the moment not the gains from solving these problems, virtually all those issues are fairly easily solveable.

Why bother solving them? Because (just as in Cell), these types operations are suited to particular special hardware. The simple fact is (as both Cell APUs and GPUs show) that its easier to scale n special CPU's (SPU or GPU ALU) than general purpose CPUs. We are looking at systems in a couple of years capable of a 1000 (real) FLOPs per cycle! Thats worth trying to adapt certain algorithms to use that power.

GPU and Cell are solutions to the computation problem, graphics were the first things that needed it, sound and physics second, AI will be third. Being able to switch where we compute things, allows us to overcome bottlenecks and produce higher quality games.

 

IGN David Roman Interview:

Source: IGN

 

Male lady bugs will beat you up for calling them that, so try not to, unless you're a slave to pain...

 

Hmmm, so MS invented PS/VS before everybody else. Nevermind the initial 3DLabs OpenGL 2.0 SL proposal and the Stanford shading language - both predate the MS HLSL spec.

VS/PS in OpenGL and DirectX are nearly identical apart from some small semantical differeneces. There's really nothing DirectX/OpenGL specific about each of the specs.

Embedded OpenGL will be based on OpenGL 2.0 minus some things not needed in a console. The MS API will be based on DX10. Is DX10 mature?

The shading languages are nearly identical. I doubt having to learn slightly different syntax will hinder the developer's abilities to write shaders.

Yeah, whatever...

 

In some ways I kind of expected this, though a part of me is a bit disappointed. I guess we don't really know what nvidia's next GPU will be like, but I don't think it will be nearly as software oriented an approach as I was hoping for. Still, the silver lining is that it probably means that nvidia's linux drivers between the Cell workstation and PCs will be very similar and should share a lot of common code. The linux drivers will probably end up being very very good (Which is much more than I can say for ATIs drivers at this point).

Nite_Hawk

 

Well there goes the visualizer. Now that means the ps3 will have a nvida graphics chip using there acrticiture(sp). Instead of using the visualizer artcututure (sp)

 

Please stop eating the brown acid, it kills brain cells.

What prompted this utterly random (and pretty much completely false) post? Do you even know anything about rambus memory except whatever (dis)information you've read posted by other clueless people like yourself on the web?

As for your silly talk about why eDRAM would be too small, you needn't worry; 3D renderering hardware don't store all the geometry in on-chip/board memory anyway, so your concerns are all moot.

 

Beat me to it. heh.

Sure, art assets can be transferred but we'll see how it all plays out.

Back to more on topic quesitons: having a custom, beefed up "next gen" nvXX chip bolted onto the cell architecture would require custom, probably hardware, interfacing, right? Or (vice versa) would require custom bridging for PC architecture? There could be significant problems with an efficient bridge between two architectures, right?

I guess I don't understand how you can do a "custom chip based on next gen tech" that is compatable. This keeps getting more and more interesting.