Our goal with the Cell is to be an order of magnitude faster

Simon F said:
Jaws said:
I can't see pin shadows against that coin...
:rolleyes: :rolleyes:
There are no pins on a chip. Instead there are pads to which gold wires are bonded. These in turn then go to the pins/balls of the package.

IIRC, the pads are usually located around the outside of the device.

Yeah wrong terminology, that's what I meant...

HAS ANYONE CONFIRMED IF THESE CHIPS ARE REAL OR A SICK PHOTOSHOP! :p
 
where did you get that info?
ugh, thats gay that the ps3 MIGHT have fewer spe's

From the guy that have seen the paper.

That's his speculation in regard to having fewer SPE's because the die size is 206^mm2, which is large for some $300 consoles, such as PS3.
 
It says that the CPU will be used in the PS3 and Xbox2. Does he mean the just PPC core will used in the XB2? I'm getting a bit lost here.
 
Pugger said:
It says that the CPU will be used in the PS3 and Xbox2. Does he mean the just PPC core will used in the XB2? I'm getting a bit lost here.

I just don't think David knows as much about next gen. consoles. So he probably assumed that CELL will go into both XB2 and PS3 since IBM is making MPUs for both.

Cheers
Gubbi
 
Cheers Gubbi I was getting a bit confused there. I wonder though if this means the Cores in the XB2 will also be rated at 4Ghz?
 
V3 said:
That's his speculation in regard to having fewer SPE's because the die size is 206^mm2, which is large for some $300 consoles, such as PS3.

If it comes out at 90nm, but on the cusp of 65nm they may be willing to soak that up for the time being.
 
207 \ 0.90² * 0.65² would be 107mm² on 65nm. Dont know, but I would be (irrationally) dissapointed if PS3 would have a single PU, regardless how it would perform. SPUs are nice, but 2 general purpose PEs would ensure the PS3 wont have to suffer penalties in quick and easy Code to "that other" Console.
 
Npl said:
207 \ 0.90² * 0.65² would be 107mm² on 65nm. Dont know, but I would be (irrationally) dissapointed if PS3 would have a single PU, regardless how it would perform.
Even if PS3 config is 1:8 they could pump up local sram size..;)
 
Remember the first iteration of PS2s CPU was 240mm2 @ 250nm. It was IIRC pretty quickly shrunk down (180nm?). It wouldn't be surprising if PS3 followed a similar course. Of course GS was even larger than EE due to the on-chip memory. Who knows how large the PS3 GPU will be; is it a CELL processor containing an Nvidia processor, or is it an Nvidia processor interfacing with an external CELL processor? The old patents that have been discussed here illustrate a large main CELL CPU and a corresponding "small" CELL processor containing a pixel processor within (Nvidia?) but who knows if those patents are outdated by now.
 
New "small" pic: http://mms.businesswire.com/bwapps/mediaserver/ViewMedia?mgid=38917&vid=3

http://home.businesswire.com/portal...iew&newsId=20050207005340&newsLang=en

IBM Analysis Engineer Tami Vogel holds a prototype of the new Cell microprocessor, a collaboration between engineering teams from IBM, Sony and Toshiba. Essentially a supercomputer on a chip, the Cell microprocessor is expected to transform consumer electronics and digital entertainment. (Photo: Business Wire)

Fredi
 
Then 2PE in the 65nm process is a safe bet.

That size is with each SPE having 256kb of memory instead of what we assumed before 128kb, that's double the memory on that chip than what we first discover in the patent.

I think single PE on a chip is all you'll get, if not they'll have to do hot potatoes or something similar to avoid thermal limit, and performance won't be any better. Which IMO a waste on something like PS3. Sony does have MCM patent though, so who knows.
 
Re: Our goal with the Cell is to be an order of magnitude fa

Inane_Dork said:
"Our goal with the Cell is to be an order of magnitude faster," said Lisa Su, an I.B.M. executive in charge of technology development and licenses.
10x faster than what?
A team of IBM, Sony, and Toshiba have collaborated on the development of the "Cell" microprocessor at a joint design center established in Austin, Texas, since March 2001. The device is aimed for entertainment and media applications — at speeds up to 10 times the performance of the latest PC processors.
[source: http://www.eet.com/conf/isscc/showA...GCKHSCJUMEKJVN?articleId=59301581&kc=3681 ]









(I first found that article from the link posted here: http://www.beyond3d.com/forum/viewtopic.php?t=19815 )
 
Re: Our goal with the Cell is to be an order of magnitude fa

Wunderchu said:
A team of IBM, Sony, and Toshiba have collaborated on the development of the "Cell" microprocessor at a joint design center established in Austin, Texas, since March 2001. The device is aimed for entertainment and media applications — at speeds up to 10 times the performance of the latest PC processors.
[source: http://www.eet.com/conf/isscc/showA...GCKHSCJUMEKJVN?articleId=59301581&kc=3681 ]









(I first found that article from the link posted here: http://www.beyond3d.com/forum/viewtopic.php?t=19815 )
Ah. That paraphrase makes more sense. Up to 10x faster than current PCs is reasonable, given certain standards (FLOP rating, for one).
 
Back
Top