jvd said:
...and Toshiba, and IBM... There's a lot of coverage for CELL itself.
Toshiba, Sony close to 65nm sample production
- Thread starter cybamerc
- Start date
...and Toshiba, and IBM... There's a lot of coverage for CELL itself.
overclocked
Veteran
I was just thinking about the "timeframe" and if we say that PS3 will be launched early 2006, maybe a limited number of quantites in Japan/maybe US in late 2005.
The first thing is that i hope that some work on the software has already "started" /atleast in Sony´s own team´s.
The second is this with the timeframe and the 65nm node, if we look at AMD their new FAB(36) should at this timeframe have volume production at 65nm.
I would like too clarify my above post also that i don´t think we need too spend soo much time about the manufacturing of BE/PS3 but instead more focus on what Sony @ co are doing with the massive transistorcount etz... It´s not worth anything if someone say "WOW the PS3 has xxxtrans and runs at xxxGhz", instead what´s IMHO important is " How they implented the logic, did/are they making good use of the budget they are on". Tradeoffs in logic must be done, are they cuting the right corners and soo on.
I vill think that they done the most right, learned from some flaws the PS2 has struggled with due too the learning curve. Oh well only time will tell but the ambition/researsh and tons of money points too something wonderful..
The first thing is that i hope that some work on the software has already "started" /atleast in Sony´s own team´s.
The second is this with the timeframe and the 65nm node, if we look at AMD their new FAB(36) should at this timeframe have volume production at 65nm.
I would like too clarify my above post also that i don´t think we need too spend soo much time about the manufacturing of BE/PS3 but instead more focus on what Sony @ co are doing with the massive transistorcount etz... It´s not worth anything if someone say "WOW the PS3 has xxxtrans and runs at xxxGhz", instead what´s IMHO important is " How they implented the logic, did/are they making good use of the budget they are on". Tradeoffs in logic must be done, are they cuting the right corners and soo on.
I vill think that they done the most right, learned from some flaws the PS2 has struggled with due too the learning curve. Oh well only time will tell but the ambition/researsh and tons of money points too something wonderful..
nondescript
Regular
Been away for awhile, so sorry if I'm bringing up old news...
In general, I still think Sony will be making PS3 chips starting from day one of mass production - there is no mention of PSP chips at 65nm, and while it makes a lot of sense that PSP production will eventually go to 65 nm (power savings), I think PS3 will still be their top priority. Besides, their other fabs still need something to do... As to when they launch, who knows. They can launch in Winter 2005 if they want to, but when they actually do will depend on what MS and Nintendo are up to (as well as their sales of PSP and PS2)
V3, thanks for the article, good to see that STI are on schedule, doing what the said they would when they said they would. That angled ion-impantation scheme sounds interesting, when I get back to school I'll get the IEDM proceedings from my advisor. It seems to me that adding a few more mask steps would do the job, but maybe this is more cost-effective.
Panajev - SSOI seems pretty experimental at this stage. I have every confidence in their methods, but strained silicon will have more crystal defects. I don't think STI will do SSOI and 65nm - 65 nm SOI is hard enough! I also don't think the method described can be easily ramped up to the volumes they need. In a nutshell, they lay down SiGe for the lattice structure, grow the strained silicon, bond it to an oxidized wafer, then introduce hydrogen, so they can cleave the wafer behind the strained Si - that's one hell of a job. Compare that with the standard SOI making procedure - oxygen implantation, then annealing. And it doesn't look like they have the quality control needed for industry applications yet.
Seriously, do you want PS3 to fail? Of course it could fail, just as I could get stabbed to death tomorrow by a flock of crazed woodpeckers. But given the fact that PS1 and PS2 have been successes, and that PS3 R&D and production seems to be moving right along, doesn't that suggest that it is much more likely that PS3 will succeed than fail?
In general, I still think Sony will be making PS3 chips starting from day one of mass production - there is no mention of PSP chips at 65nm, and while it makes a lot of sense that PSP production will eventually go to 65 nm (power savings), I think PS3 will still be their top priority. Besides, their other fabs still need something to do... As to when they launch, who knows. They can launch in Winter 2005 if they want to, but when they actually do will depend on what MS and Nintendo are up to (as well as their sales of PSP and PS2)
V3, thanks for the article, good to see that STI are on schedule, doing what the said they would when they said they would. That angled ion-impantation scheme sounds interesting, when I get back to school I'll get the IEDM proceedings from my advisor. It seems to me that adding a few more mask steps would do the job, but maybe this is more cost-effective.
Panajev - SSOI seems pretty experimental at this stage. I have every confidence in their methods, but strained silicon will have more crystal defects. I don't think STI will do SSOI and 65nm - 65 nm SOI is hard enough! I also don't think the method described can be easily ramped up to the volumes they need. In a nutshell, they lay down SiGe for the lattice structure, grow the strained silicon, bond it to an oxidized wafer, then introduce hydrogen, so they can cleave the wafer behind the strained Si - that's one hell of a job. Compare that with the standard SOI making procedure - oxygen implantation, then annealing. And it doesn't look like they have the quality control needed for industry applications yet.
jvd said:
Seriously, do you want PS3 to fail? Of course it could fail, just as I could get stabbed to death tomorrow by a flock of crazed woodpeckers. But given the fact that PS1 and PS2 have been successes, and that PS3 R&D and production seems to be moving right along, doesn't that suggest that it is much more likely that PS3 will succeed than fail?
Yes its still valid even if everyone in the world was working towards cell itself .[/quote]jvd said:
Ok, now that's just rediculous. I mean, you're including John Romero, and he'd fuck up ANYthing!
cthellis42 said:
Ok, now that's just rediculous. I mean, you're including John Romero, and he'd fuck up ANYthing!
[/quote] haha . I'm just saying that just because there is alot of money pumped into cell does not mean it wont fail. It can reduce the likely hood that it would fail but only so much.Vince said:jvd said:
Yes, it's still prone to all those "Act of God" events (Terrential flooding, Astroid Impacts, Thermonuclear War) - you know, the kinds of things which should be repeated in every post after you make a positive comment on STI.
You know using your logic I can say ms is going to pump millions if not billions into xbox2 so of course it can't fail. But as we see with xbox 1 money doesn't mean anything in the end.
And this was not in anyway a negative comment to sti. This is just a fact. Because I do not view sti as a god that can not make a mistake is not a negative view on sti nor comment. It is simply just a smart observation .
overclocked
Veteran
Jvd, this has nothing at all too do with Duke Nukem Forever...
jvd said:
Not really, since that's taking into account all kinds of market pressures that they have no control over. If, however, they were spending four years and billions of dollars on one machine--one chip--I would expect there to be a VERY good chance that it would kick ass. Hence why I referred to CELL itself--as a chip. As an architecture.
Its challenge will be "what does it mean?" The software side. How much can they tap, and what will it translate to, and how well can they get others to do it? With the PS3's it's in what implementation of CELL Sony can squeeze into their box and just how it translates to a gaming experience all around. (And that in itself will be affected by a great many non-CELL matters as well.)
The phrase "anything can fail" is a no-brainer, but rather useless to mention as it applies universally to everything and in the end says nothing. Do I think CELL itself is going to be a dud as an architecture, deliver a tiny fraction of what they're talking about, fail to work, suffer horrible yields, get delayed for a year...? Certainly doesn't look like it so far. I'd be plenty more interested in Xbox 2 if Microsoft were working on it years ago, spending far more resources, adopting partners early, and having a clear goal from the beginning--and even if they were adopting a whole new plan, unless there were word coming out about setbacks I'd be pretty confident in their delivery. (I'm interested enough in X2 as it is, but for entirely different reasons.)
As to what the PS3 will "be" or "mean" or "do"... who the hell knows? I think it probably WOULD take an act of god to make it "suck" overall--I don't think I know ANY consoles that I would ever say that about--and lord knows that unless it's a pile of shit it will certainly sell well. (Too much momentum and market trends we can already see for it to not.) But just where it starts, how "good" it will be, how it will stack up to the competition, how far it will progress...? <shrugs> Finding out is the most interesting part. ^_^
Certainly the back-and-forths we have now are hardly going to change form no matter WHAT it looks like!
JVD, in fact my post was not referred to MONEY alone, in fact that was one of the weakest elements in the potential success of Cell.
I stated that the combination of:
- years of development
- IBM-Tosh-Sony alliance, and subsequently their own expertise put into one single project
- money
All push this project into the realm of "likelihood of success". Then OF COURSE everything can fail. Nvidia has gone way down when they were at the peak of their success (although that has individual variables into it which are not present in this case, see: NV2A sucked their resources).
Anything can fail, anything can happen, here we're discussing about an average case scenario where meteorites don't fall on Sony's fabs, earthquakes don't kill every MS employee and such.
If MS had been pumping billions upon billions on a new project with 2 other super-powers of electronics, then OF COURSE that would push them into the green area of "likelihood to succeed"...
However, only money can't do this alone, and that is why is said the above elements are more important than money itself... MS could spend a trillion on a new chipset developed with S3, it would still be crap... With all due respect for S3, just giving an example..
It's common sense really...
I stated that the combination of:
- years of development
- IBM-Tosh-Sony alliance, and subsequently their own expertise put into one single project
- money
All push this project into the realm of "likelihood of success". Then OF COURSE everything can fail. Nvidia has gone way down when they were at the peak of their success (although that has individual variables into it which are not present in this case, see: NV2A sucked their resources).
Anything can fail, anything can happen, here we're discussing about an average case scenario where meteorites don't fall on Sony's fabs, earthquakes don't kill every MS employee and such.
If MS had been pumping billions upon billions on a new project with 2 other super-powers of electronics, then OF COURSE that would push them into the green area of "likelihood to succeed"...
However, only money can't do this alone, and that is why is said the above elements are more important than money itself... MS could spend a trillion on a new chipset developed with S3, it would still be crap... With all due respect for S3, just giving an example..
It's common sense really...
Talking about expectation and failure, my expectation for PS3 is what I read in that patent. Anything less, and I'll be dissapointed.
For Xbox2, I expect some MCM with 4 dual core PPC chips, with massive L2 and L3 cache like greater than 128 MB, dual R500 variant chip, huge memory like 2-4 GB, with bandwidth > 100 GB/s, huge HD at least 200 GB. All by 2005 for $300. Considering what they had this generation, anything less than this I'll be dissapointed.
For N5, I'll be dissapointed if they don't launch with a new Mario game like they did this generation.
For Xbox2, I expect some MCM with 4 dual core PPC chips, with massive L2 and L3 cache like greater than 128 MB, dual R500 variant chip, huge memory like 2-4 GB, with bandwidth > 100 GB/s, huge HD at least 200 GB. All by 2005 for $300. Considering what they had this generation, anything less than this I'll be dissapointed.
For N5, I'll be dissapointed if they don't launch with a new Mario game like they did this generation.
With specs like that V3 we could have a GTA type of game with characters that look like the fairy in the Nvidia Dawn demo and cars that look like the "rusty car demo" (can't remember the name)
2-4GB Ram seems a bit much even in 2006. High end PCs might have that much RAM those days, i'm somewhat skeptical about a console having that much memory.
And double R500+4Core PPC??? Don't know... It all depends when this this is going to be released..
Still, what do we know...
2-4GB Ram seems a bit much even in 2006. High end PCs might have that much RAM those days, i'm somewhat skeptical about a console having that much memory.
And double R500+4Core PPC??? Don't know... It all depends when this this is going to be released..
Still, what do we know...
If R500, is what I expected it to be, you probably get better looking characters than that fairy, and you can have the environment that surpasses Half life 2 all in high res too.
Crap man, I got that much on my current PCs. Memory are like really cheap now days, and its getting faster too. Highend graphics card has like 256 MB now, next year, they'll probably have 512 MB - 1 GB version.
Might as well, if Sony can put 32 processors, what's stopping MS from putting 4 dual or even quad cores PPC chip, giving like 16 cores PPC.
zidane1strife
Banned
significantly surpass...
Panajev2001a
Veteran
V3, 1 APU should be definately smaller than a normal PowerPC 970 ( the G5 is a full blown CPU designed for Desktop PCs ), so we cannot really work 32 for Sony, 16 for MS as your 16 are full sized PowerPC cores.
Similar threads
