• Hey Guest. Check out your NeoGAF Wrapped 2025 results here!

Kutaragi talks about Cell history

Andrew2

Banned
A little snippet from the interview:

When did you yourself start forging the Cell initiative?

I started thinking of the idea before the launch of the PlayStation 2 (PS2), around the summer of 1999. There were a mound of things needing to be done about the PS2 in terms of nurturing it as a business, but in terms of technological development, it was already a thing of the past to me.


full interview

http://techon.nikkeibp.co.jp/english/NEWS_EN/20050407/103542/
 
kutaragi2.jpg


"Not every engineer has the chance to participate in such projects that can impact the course of our history. Cell is exactly this kind of project. In the day-to-day world, engineers design microprocessors with model numbers. In this project, however, you are to create a chip that has a proper name. You never know, thirty years from now, some members of our team may end up being looked up to just like Shockley or Moore."
 
To be quite honest, his ambitions might be very big, but you can't discredit him for trying to be a forward-thinking visionary. When it comes down to technology and the video game business, its a rare sight to see someone with a grande vision and Kutaragi seems to be one of those rare sights. If I was to say anything, a visionary like him the gaming industry really needs.
 
The network of today still remains a means to access information. The time will come, however, when the network itself becomes a computer. So far, we have pursued how to improve the performance of a single chip, but we are now becoming aware of the limits that loom there. Once you suppose that the network is a computer, what you have to do is to improve the computing performance of the network as a whole. This is s paradigm shift. Be it on-chip, on-board, or on-network, chips will be infinitely connected -- and that is what this "Cell" is all about.

:lol
 
^^^ I would very much tend agree with him as Information Technology is a area of study for me. In the end though, despite having bold ambitions, you have to respect him for being visionary.
 
I wonder if CELL will actually ever become anything significant. So many things sound nice on paper but can't be put to realistic use in the real world.
 
part 2

And you chose Toshiba and IBM as your partners.

I was rather optimistic in getting support out of Toshiba, as I knew a lot of people there through the PS2 project. In addition to Toshiba, I wanted IBM to be a part of this project, as this project was all about setting off a paradigm shift. Some say I went after IBM because they had the SOI technology, or because they had PowerPC technology, but these weren't the reasons. These two technologies are undoubtedly important in realizing Cell, but what mattered most to me was to work with a partner that could really share with us the strong determination to change computing history. We all know that IBM is the leader of the computer industry. The company created the era of the mainframe with System/360. As the word "PC" once almost meant exclusively "IBM PC," PC technologies originated from IBM technologies. In other words, IBM had always been there in the history of computers. And so, IBM was indispensable for this project to create the next computing paradigm.



part 3

I intend to stick to 90nm technology. I know some people who suggested going into mass production using such outrageous technologies as 65nm or 45nm, but you can't make something only by dreaming. In my management role, I'm determined to start mass production with the 90nm.
 
Everyone that thinks Cell in PS3 will have 6SPE and not 8 should read this:

Cell has 8 embedded "SPE" CPU cores. What is the basis for this number?

Because it's a power of two, that's all there is to it. It's an aesthetic. In the world of computers, the power of two is the fundamental principle - there's no other way. Actually, in the course of development, there's this one occasion when we had an all-night, intense discussion in a U.S. hotel. The IBM team proposed to make it six. But my answer was simple - "the power of two." As a result of insisting on this aesthetic, the chip size ended up being 221mm2, which actually was not desirable for manufacturing.

In terms of the one-shot exposure area, a size under 185 mm2 was preferable. I knew being oversized meant twice the labor, but I on the other hand, I thought these problems of chip size and costs would eventually be cleared as we go along. But in this challenge of changing the history of computing, I could not possibly accept any deviation from the rule of the power of two. For example, the world of communication also has gone to the rule of the power of ten. Ethernet, which started with 10M bit/s, has gone through stages of 100M and 1G, and 10G is certain to come next. You won't go with, say, 4G just because 10G is technologically difficult. It is my belief that real technological innovation is born from such persistence.

Cell in PS3 will have 8 SPE, Kutaragi would never allow a 6SPE Cell in the PS3.

Fredi
 
McFly said:
Everyone that thinks Cell in PS3 will have 6SPE and not 8 should read this:



Cell in PS3 will have 8 SPE, Kutaragi would never allow a 6SPE Cell in the PS3.

Fredi


Gotta love how Kutaragi is so adament about things like this.
 
McFly said:
Everyone that thinks Cell in PS3 will have 6SPE and not 8 should read this:



Cell in PS3 will have 8 SPE, Kutaragi would never allow a 6SPE Cell in the PS3.

Fredi


But he got demoted recently; does he still have enough power to allow this?
 
MassiveAttack said:
Is anyone seriously betting AGAINST Kutaragi next gen - i.e. actually expecting him to fail to maintain Sony's leadership?

Step up.


Hardware wise, no. Kutaragi is unique because he's so damn focused and driven, he gives Sony an edge which I think Nintendo lost after about 1990. Call it the "eye of the tiger" or whatever you want, he's got it.

Kutaragi still runs the game division, he calls the shots as far as PS3 is concerned.

Maybe though its this "ruthless" ambition/vision he has that was one reason why Sony opted to go with a more "diplomatic" suit as their president (Stringer) though.

Just because Michael Jordan is you best basketball player, doesn't neccessarily mean you want him playing the point guard position for instance.
 
McFly said:
Everyone that thinks Cell in PS3 will have 6SPE and not 8 should read this:



Cell in PS3 will have 8 SPE, Kutaragi would never allow a 6SPE Cell in the PS3.

Fredi

His comments are not necessarily at odds with the concept of a PS3 CPU with fewer than 8 useable SPEs. The chip can be manufactured with 8 SPEs but not have 8 exposed to the programmer..

Also, interesting comment on yield and costs:

At this scale, brute strength alone won't do. We are not aiming to produce 100,000-yen chips. As Mr. Masatoshi Shima, who had designed the "4004" chip with Intel once wrote, Intel usually starts with the yield of 40% and then brings it up to 85%. That is a good curve.

Doom_Bringer said:
But he got demoted recently; does he still have enough power to allow this?

He's still in charge of playstation, and he'll decide what the CPU in PS3 looks like. But see above - his comments don't necessarily contradict a different number of useable SPEs in PS3's CPU.
 
soundwave05 said:
Hardware wise, no. Kutaragi is unique because he's so damn focused and driven, he gives Sony an edge which I think Nintendo lost after about 1990.

Kutaragi still runs the game division, he calls the shots as far as PS3 is concerned.


cool! I hope it has 8 SPEs :P
 
Doom_Bringer said:
But he got demoted recently; does he still have enough power to allow this?

Demoted from electronics and semiconductor business,not from SCEI.Every SCE's product must have his approval.
 
gofreak said:
He's still in charge of playstation, and he'll decide what the CPU in PS3 looks like. But see above - his comments don't necessarily contradict a different number of useable SPEs in PS3's CPU.

Same as my reply in Cell thread:
You're right to say it doesn't mean 6 SPE's are ruled out. My opinion (FWIW) is that Sony's intention at this stage is to try and produce chips with 8 workable SPE's. However if the yield is too low with 8 workable, might Ken go for 6? If I was him I would. And might that have informed his decisions on SPE numbers in the first place? Probably.
 
gofreak said:
His comments are not necessarily at odds with the concept of a PS3 CPU with fewer than 8 useable SPEs. The chip can be manufactured with 8 SPEs but not have 8 exposed to the programmer..
Based on that passage in the interview, would there really be a difference in Ken's mind? Either way, it fails his "power of 2" stance. I don't think it's merely manufacturing process to which he wants to employ the power of 2 - this all started back during the design of the chip, according to him. Sticking to his rule, the only way they might use Cell chips that don't have all 8 SPEs available is if it were 2 or 4 usable SPEs...would that level of manufacturing defect be considered acceptable?
 
gmoran said:
My opinion (FWIW) is that Sony's intention at this stage is to try and produce chips with 8 workable SPE's. However if the yield is too low with 8 workable, might Ken go for 6? If I was him I would. And might that have informed his decisions on SPE numbers in the first place? Probably.

Something tells me these scenarios have already played out to a certain degree (though not quite exactly as you describe..) ;)

kaching said:
Based on that passage in the interview, would there really be a difference in Ken's mind? Either way, it fails his "power of 2" stance. I don't think it's merely manufacturing process to which he wants to employ the power of 2

I dunno, he talks a lot about manufacturing in that comment..
 
McFly said:
Everyone that thinks Cell in PS3 will have 6SPE and not 8 should read this:



Cell in PS3 will have 8 SPE, Kutaragi would never allow a 6SPE Cell in the PS3.

Fredi

If someone, who is not me, suggested that the final chip might look like 6 and at the same time as 8 or maybe also in another way...

What an insane guy :lol.

Now, spin around that one :lol.
 
kaching said:
Based on that passage in the interview, would there really be a difference in Ken's mind? Either way, it fails his "power of 2" stance. I don't think it's merely manufacturing process to which he wants to employ the power of 2 - this all started back during the design of the chip, according to him. Sticking to his rule, the only way they might use Cell chips that don't have all 8 SPEs available is if it were 2 or 4 usable SPEs...would that level of manufacturing defect be considered acceptable?

If four or even six SPE's are defect on most chips, then the odds are high that the PE is defect as well. With a yield like that you could never go mass producing.

Fredi
 
Panajev2001a said:
If someone, who is not me, suggested that the final chip might look like this: it will have 6 SPE's, but it will have 8 SPE's.

Now, spin around that one :D.

Bad Panajev :p
 
McFly said:
If four or even six SPE's are defect on most chips, then the odds are high that the PE is defect as well. With a yield like that you could never go mass producing.

Fredi

I think the PPE should not be the most problematic piece yields wise: it is a well tested design which will go under even more stress in Xenon's CPU manufacturing (a more complex version of that core).

I think they are more worried about SPE's redundancy than PPE's manufacturing problems.
 
As a result of insisting on this aesthetic, the chip size ended up being 221mm2, which actually was not desirable for manufacturing

I'd forgotten he'd said this as well: so I don't think he's intention was ever that scrubbing SPE's might help with yields. Whats the point of adding extra SPE's to help with yields if there are other complications in manufacturing in adding them in the first place. Doesn't seem reasonable. However now he's got them, he can sacrifice them if needed, I'd imagine that would be last resort type stuff though.
 
Panajev2001a said:
Either he changes the tune for PlayStation 3's CPU and/or GPU or I owe nAo/Nostromo a meal... and the restaurant ain't even cheap... :lol.
Mmmmhhhh..I'm so hungry!
(italian) food is even better when it's free.. :lol
 
Wow, that was interesting and I'm not much of a tech person. Kutaragi is much more hardcore engineer, almost religously so, than I knew from passing interviews in general magazines and such. Very cool.
 
gofreak said:
I dunno, he talks a lot about manufacturing in that comment..
Sure, he does, but as gmoran pointed out, he makes the comment that he understood that going to 8 SPEs would make the chip size uncomfortably large for manufacturing. He then comments on ways to decrease chip size over time which made him think this would be an acceptable disadvantage in the short-term.
 
If you have even the passing interest in this kind of technologies, you've got to respect this guy. It's not everyday that you see someone who puts visions first and financials second, to make something he would be proud of, and other people will like :) Sadly, I kinda have a feeling that way of thinking used to be a lot more common some decades ago than it is today.
 
goomba said:

goomba apparently doesn't understand GRID computing and what Internet 2 is all about.

Don't stand out in the sun when your head is made of wax.

Cell has much bigger plans behind it than just the piddly PS3. And Cell's architecture is pretty damn amazing. Multi-core CPUs from Intel and AMD are archaic compared to it.
 
Panajev2001a said:
If someone, who is not me, suggested that the final chip might look like 6 and at the same time as 8 or maybe also in another way...

6 programmable, 2 providing some kind of automated support, or hidden in the background? But why? to ease programming? Why wouldn't you expose all available resource to a developer?
 
mrklaw said:
6 programmable, 2 providing some kind of automated support, or hidden in the background? But why? to ease programming? Why wouldn't you expose all available resource to a developer?

Uhm... close, but no cigar...
 
Panajev2001a said:
FUCK, FUCK, FUCK, FUCK!!!!

Either he changes the tune for PlayStation 3's CPU and/or GPU or I owe nAo/Nostromo a meal... and the restaurant ain't even cheap... :lol.



hmmmm what was the bet ? assuming there was '__'


umm so I guess that we will look forward to.... PS3 starting out on 90nm (unless a miracle happens) then soon migrate to 65nm and then onto 45nm later in its lifecycle.

then PS4 debuts on a sub-45nm process (35nm 32nm 25nm or whatever) and will have the equivalent of 100 or more SPE-2 processors providing teraflops of processing power in a console (sounds like old PS3 talk, lol)
 
Just think of the amount of chips we are to produce - 20 million for game consoles alone

20 million for PS3? Not lifetime, so when? 20m for the 90nm process? 20million for year one?
 
mrklaw said:
Please Make things clearer then....

Never used that before, so screw it :)

It will be mentioned soon I think, just few more weeks... in the meanwhile they might still go to 2 SPE's :lol.
 
slight correction: unless i am mistaken, it seems process technology, process nodes, etc will migrate from 90nm to 65nm to 45nm to 32nm to 22nm to 16nm <---- PS4 on 22nm or 16nm ?
 
midnightguy said:
hmmmm what was the bet ? assuming there was '__'


umm so I guess that we will look forward to.... PS3 starting out on 90nm (unless a miracle happens) then soon migrate to 65nm and then onto 45nm later in its lifecycle.

then PS4 debuts on a sub-45nm process (35nm 32nm 25nm or whatever) and will have the equivalent of 100 or more SPE-2 processors providing teraflops of processing power in a console (sounds like old PS3 talk, lol)

I cannot really comment on PlayStation 4, too far off.

PlayStation 3 seems indeed launching on 90 nm technology and then moving to 65 nm and then 45 nm which for Sony might be ideal if they can have better yields with a better known manufacturing process.
 
If someone, who is not me, suggested that the final chip might look like 6 and at the same time as 8 or maybe also in another way...
And what it someone (who is not me) told you that there's already a chip that looks like that out there, and is by no means final. :P
 
Fafalada said:
And what it someone (who is not me) told you that there's already a chip that looks like that out there, and is by no means final. :P

I would say, gee-whiz captain... you rok my boat ;).
 
Fafalada said:
And what it someone (who is not me) told you that there's already a chip that looks like that out there, and is by no means final. :P

I guess we'll probably know in about 2+ weeks or so - unless real info leaks out early, hehe :)
 
Top Bottom