HotChips 17 - More info upcoming on CELL & PS3

Status
Not open for further replies.
D

dcforest

Newcomer
I don’t recall seeing this information posted, but if it’s a duplicate, I apologize in advance.

The upcoming annual HotChips conference (Aug. 14-16th) should have some information about CELL & PS3, as there is an entire session dedicated to CELL.

9:30-11:30am Session One: Cell Processor
Session Chair: Mitsuo Saito

Presentations:
• A Novel SIMD Architecture for the CELL Heterogeneous Chip-Multiprocessor, IBM
Authors(s): Michael Gschwind, Peter Hofstee, Brian Flachs, Marty Hopkins, Yukio Watanabe, Takeshi Yamazaki

• IBM CELL Interconnect Unit, Bus and Memory Controller, IBM
Authors(s): Scott Clark, Kent Haselhorst, Kerry Imming, John Irish, Dave Krolak, Tolga Ozguner

• Super Companion Chip w/ A/V oriented interface for the CELL Processor, Toshiba
Authors(s): Takayuki Mihara, Kenichi Ishii, Naoki Sugawa

• Programming and Performance Evaluation of the CELL Processor, Toshiba
Authors(s): Ryuji Sakai, Seiji Maeda, Christopher Crookes, Mitsuru Shimbayashi, Katsuhisa Yano, Tadashi Nakatani, Hirokuni Yano, Shigehiro Asano, Masaya Kato, Hiroshi Nozue, Tatsunori Kanai, Tomofumi Shimada, Koichi Awazu

Please note the Toshiba paper: Super Companion Chip w/ A/V oriented interface for the CELL Processor. There is more to PS3 than STI has disclosed so far. ;)

Also, day two will have some good info too as David Kirk is giving a presentation and he is always informative.

1:20-2:20pm Keynote II
Keynote Chair: John Nickolls

Presentation:
• Multiple Cores, Multiple Pipes, Multiple Threads - Do we have more Parallelism than we can handle?
Authors(s): David Kirk (Chief Scientist, NVIDIA)

Microsoft has a undisclosed processor presentation coming as well on the second day (see bottom presentation).

4:20-6:20pm Session Eight: Processors and Systems
Session Chair: Mark Tremblay

Presentations:
• Intel 8xx series and Paxville Xeon-MP Microprocessors, Intel
Authors(s): Jonathan Douglas

• TwinCastle: A Multiprocessor North Bridge Server Chipset, Intel
Authors(s): Debendra Das Sharma, Ashish Gupta, Gordon Kurpanek, Dean Mulla, Bob Pflederer, Ram Rajamani

• Dynamically Optimized Power Efficiency with Foxton Technology, Intel
Authors(s): Samuel Naffziger

• Presentation To Be Announced, Microsoft
 
Apologies dc for not looking it up myself - but an Sony speakers there? KK would be a blast! - but no names I recognise.
And I'd always assumed that Hotchips was more of a semiconductor type event rather than a "finished article" thing.
 
dcforest said:
Please note the Toshiba paper: Super Companion Chip w/ A/V oriented interface for the CELL Processor. There is more to PS3 than STI has disclosed so far. ;)
Click to expand...

Hmm..something you're speculating or something you know...?
 
Surely that's just a chip Toshiba are using for audio/video work on their own CE goods? Why is this considered part of PS3, unless the whole first day is PS3?

My guess is it's a derivative of the Cell graphics part intended for PS3 before nVidia won the deal and has nothing to do with PS3.
 
slider said:
Apologies dc for not looking it up myself - but an Sony speakers there? KK would be a blast! - but no names I recognise.
And I'd always assumed that Hotchips was more of a semiconductor type event rather than a "finished article" thing.
Click to expand...
T. Yamazaki is from SCE (Check out here)
 
Shifty Geezer said:
Surely that's just a chip Toshiba are using for audio/video work on their own CE goods? Why is this considered part of PS3, unless the whole first day is PS3?

My guess is it's a derivative of the Cell graphics part intended for PS3 before nVidia won the deal and has nothing to do with PS3.
Click to expand...

This would seem the likely situation, but the wording is a little odd.

"Super Companion Chip w/ A/V oriented interface for the CELL Processor"

I guess one might read it as a Cell Chip with a A/V orientated interface i.e. a "super companion chip". I'm guessing "w/" is "with"?
 
Titanio said:
dcforest said:
Please note the Toshiba paper: Super Companion Chip w/ A/V oriented interface for the CELL Processor. There is more to PS3 than STI has disclosed so far. ;)
Click to expand...

Hmm..something you're speculating or something you know...?
Click to expand...

Really? A thrid chip?
 
Since I didn't know where else this should go..

Probably a misunderstanding but:

According to Zeitler, early successes have also been seen the likes of Sony, Microsoft, and Nintendo sign up to collaborate on Cell developments for future versions of their games machines. "Game companies are the first evidence of a move from microelectronics to introduce technology like Cell into their products," he said.

"You'll see this in a whole variety of industry segments," he added. "I think with Cell, any application that does any sort of image processing you'll see an advantage." The major difference with Cell, according to Zeitler, is that partners get to collaborate on development, rather than being presented with a finished processor.
Click to expand...

From: http://www.cbronline.com/article_news.asp?guid=98A4A3D1-C6B3-42BF-81EF-6C5FF2EA94EE

Very probably a misunderstanding, I suppose :?
 
IIRC, I think it was at HotChips in 1999 where Ken Kutaragi first talked about the roadmap from EE/G2 to EE2/GS2 to EE3/GS3 and beating down Intel with process technology and transistor counts.
 
For Sony to license Cell to its competitors would be suicide! They make money from software, not hardware, and enabling the hardware (if Cell proves that good) of their rivals would surely cut into Sony's software base.

I'd like to know in what context that statement is made - when and where, and were they talking about Cell exactly, or just 'Cellular architectures' (read multicore)?
 
Well, if nothing else it's just another example of an article basically calling the Cell processor strictly IBM's creation - and oh these other guys had some input now and then...

I also recall wondering before if IBM's move to offer customized versions of Cell would open it up to future use by Microsoft (or Nintendo); a question that began in my head after that EETimes open-source article back in the day. Now, I'm thinking that this article probably got it wrong, if only because we'd be hearing about this everywhere otherwise (and the way the rumor-mills work, maybe today we will). I still wonder though if Sony has some sort of non-compete clause worked into the technology agreement with IBM and Toshiba. They certainly should. Though I could create a scenario in my head in which Sony might actually want Nintendo and Microsoft using Cell derivatives, the whole thing just seems a little too off in the future for those companies to already be in talks - and again I wonder if they would even have the option.
 
Is anybody attending the Hotchips conference at Stanford? I'm wanting to attend. It looks like the price of attending for students is pretty cheap. If anyone is going, care to split a hotel room? Or is there anyone that lives in the stanford area that wouldn't mind a visiter for a couple nights? :)

AR
 
There's no webcast or anything for the conference right? The Cell portion should begin in an hour. I get failing the Net, we'll just have to endure the slower flow of information.
 
"Multiple Cores, Multiple Pipes, Multiple Threads - Do we have more Parallelism than we can handle?
Authors(s): David Kirk (Chief Scientist, NVIDIA)"

That could be an interesting presentation in light of recent discussion ;)
 
Status
Not open for further replies.

Similar threads

S
  • Locked
Kutaragi new Interview?
Replies
3
Views
3K
Neeyik
N
H
IBM backstabbing sony and neglected Apple ? (long read)
5 6 7
Replies
126
Views
47K
Fox5
F
A
Full transcript of John Carmack's QuakeCon 2005 Keynote
3 4 5
Replies
98
Views
34K
nemisis
N
O
The Engineers Who Created Cell
2 3
Replies
55
Views
14K
Carl B
C
P
Interesting GDC 2005 topics :).
Replies
11
Views
4K
Wunderchu
W
Back
Top