Details trickle out on CELL processor...

[Xenus]

I think you missed the part where your supposed to be impressed by the specs Qroach. 8)

 

[Jov]

We still don't know how many Cells are there in the Workstation to give us an indication of the capabilities per Cell.

 

[Qroach]

They have been sampling 65nm (confirmed) and I heard it's been since summer. 65nm production has been stated as 1H2005. Try not to invent bad things...

relax vince I'm just asking questions. I'm just trying to make sense of the hype I'm reading, such as...

"IBM plans to begin pilot production of Cell microprocessors at its 300mm wafer fabrication facility in East Fishkill, NY during the first half of 2005."

http://www.sony.net/SonyInfo/News/Press/200411/04-1129E/

Can you point me to where you found it was "confirmed"?

 

[Qroach]

Xenus,

I think you missed the part where your supposed to be impressed by the specs Qroach.

No I fell for that with PS2 last time. I'll wait for performance numbers on something that isn't strictly theoretical.

 

[Vince]

Google is your friend.

InfoWorld.[/url] 7.12.2004]Toshiba is slightly further along in the development of the technology and is currently evaluating early 65-nanometer samples, said Junichi Nagaki, a company spokesman. Toshiba is using a pilot line at a facility in Yokohama to produce the chips.

One of the first uses for Toshiba's technology will be the production of the Cell processor, which will be used in Sony Computer Entertainment Inc.'s (SCEI) upcoming PlayStation 3 games console and future consumer electronics products from other companies. In this area Toshiba is working with SCEI, Sony Corp. and IBM Corp. on technology development. Mass production is scheduled for the first half of 2005, said Nagaki.

 

[Joe DeFuria]

Google is your friend.

To be clear, a pilot line producing samples is not nearly the same thing as a production line generating production samples. In other words, I don't see the two statements (about the status at IBM's plant Toshiba's pilot plant) being contradictory.

 

[j^aws]

I love this:

Multi-thread, multicore architecture.

Now I would like to know if this multithread support thing is just an OS thing..or something more!

Hehehe...multi-threading APUs? ...We still don't know the make-up of the FPUs/FXUs within the APUs...FMADDs/FMACs or something else?

The 294.4 GFlops per PE is still based on 8 ops per cycle per APU @ 4.6 Ghz, 90 nm.

Also is it me or is the 4.8 Ghz SRAM quoted by EETimes does not match the 4.6 GHz PE speed? :?

 

[Qroach]

Joe, am I missing something? who said they were contradictory?

 

[Vince]

To be clear, a pilot line producing samples is not nearly the same thing as a production line generating production samples. In other words, I don't see the two statements (about the status at IBM's plant Toshiba's pilot plant) being contradictory.

Um, when Toshiba is evaluating 65nm sample today (err.. back in July) we assume it's from a pilot line. We also assume that when Toshiba states that "Mass production is scheduled for the first half of 2005," we assume they mean mass production.

Mass production... it does fit in with the comments of 15,000 wafers/month capacity stated for 2005. I told you, process techology is make or break.

 

[Qroach]

oh never mind joe. Got it.

 

[j^aws]

Look how many chips on a rack there are (64 chips), each chip having 32 cells on them?

There are smaller rack you know.

How about something like a 16 'blade' GScube for 16TFlop WS, 1Tflop per blade>PS3?

 

[j^aws]

Sony said it would launch home servers and high-definition televisions powered by Cell in 2006, and reiterated plans to use the microchip to power the next-generation PlayStation game console, a working version of which will be unveiled in May.

HomeServer= PSX2(<--silly name) based on PS3?

 

[mech]

So is it 4.6Ghz then? Does that mean I was right way back when?

 

[j^aws]

http://finance.yahoo.com/mp#rmbs

10:23AM Mike Tarsala's TechWatch Alert -- RMBS (RMBS) 24.34 +2.17: Analysts we reached this morning say that Rambus is very likely to benefit from the Cell processor announced today by Sony (SNE), IBM (IBM) and Toshiba (TOSBF). Of note, Rambus announced back in January 2003 that it had licensed its XDR memory and its Redwood high-speed parallel interface (its Yellowstone and Redwood technologies) to Sony and Toshiba. While IBM was NOT MENTIONED in that 2003 release, Big Blue is known for having a strict approval process when its name appears in other company's press releases. Another company analysts have mentioned as possible benefactor from the Cell processor, which is expected to go toe-to-toe with Intel in the home entertainment (specifically the multimedia living room) could be graphics-processor maker Nvidia (NVDA) -- although none of the analysts offered any specific proof. More details on the Cell processor are expected to be announced at the International Solid State Circuits Conference, which is scheduled to begin on Feb. 6 in San Francisco.

PS3 = POWER CPU + Nvidia GPU + Linux OS + OpenGL

XB2= POWER CPU + ATI GPU + Windows OS + WGF

The intriguing thing is that the ISSCC 2005 paper reports so far don't seem to mention anything about a Cell Visulaizer (aka Realizer) type GPU/Rasterizer...hmm :?

I would've thought that if Nvidia did have anything to do with it, it would be their opensource Cg shading language and compilers/drivers for the GPU/ Cell? Both Sony and Toshiba have a hatful of GPU patents they could use rather than to ask for Nvidia for hardware assistance?

 

[j^aws]

Re: Cell Performance

Hello,

Welcome to the boards Paul Zimmons! ...or is it Deadmeat!...just kidding! ...or am I...

With such small on-chip memories and such huge processing horsepower, is ray tracing or some form of GI inevitable?

If the bandwidth is there, I'd like to think realtime REYES is possible that could run Raytracing/GI enabled shaders!

 

[Joe DeFuria]

I told you, process techology is make or break.

Not to rehash an old argument, but you told me process technology is "everything", and I told you process technology is "one factor of many".

But anyway...

 

[j^aws]

I don't see what the facination is with the need for realtime ray tracing. For offline rendering, you really don't need ray tracing to have a good quality end result. Hasn't pixar proved that over and over again?

Yeah...PRman has pretty much nailed with Pixar, especially with it's raytracing/GI module... 8)

http://www.techimage.co.uk/resources/images/Teapot2.swf

 

[j^aws]

[*] 15TFlop/sec DevKits have been sent out[/list]

Thanks for summary! ...Is this the case of the missing TFlop!...I thought it was 16TFlop...also,

The companies expect that a one rack Cell processor-based workstation will reach a performance of 16 teraflops or trillions of floating point calculations per second.

I assume this is 2nd gen Cell?

As for the Cell-based workstation, it's clearly only at the prototype stage, IBM and Sony having come up with an "experimental model".

Still, it packs 2 teraflops into a standard (presumably) rackmount box, apparently, with what sounds like multiple, multi-core chips operating as a kind of cluster-in-a-box configuration. ®

http://www.theregister.com/2004/11/29/ibm_sony_cell_debut/

The Register is claiming the prototype WS is 2TFlop...

Others to add to your list of note,

• Supports multiple operating systems at the same time.
• Real-time resource management system for real-time applications.
• Multi-thread, multicore architecture.

The above article also mentions this,

More interesting is the integration of a security sub-system. The companies don't go into any detail, but it sounds not unlike VIA's PadLock technology with its hardware random number generator.

 

[Paul]

Still say a TFLOPS BE is imposible?

 

[Cryect]

I'm curious where did this 15 TFLOPs workstation come from?

And why would Sony be sending 15 TFLOP workstations for a console thats only 1/15th that at most? (heh I sure hope its not so we can have really nice CGI cutscenes)

The Register link Jaw's points to sounds definately more believable.

I have to say that rack looks nice and maybe the multi-threading is their idea of hyperthreading? Just throwing out a wild guess with that one.