View Full Version : PS3's CELL processor 'dead in the water'
According to reports, the PS3's CELL processor is 'dead in the water' and will never be used in the PS4. Apparently, the CELL line will be terminated and the current PoweXcell 8i is the last one in the series.
<blockquote>The PS3’s innovative ‘Cell’ processor won’t be used in the PS4, according to reports this morning, claiming the chip is “dead in the water”. Apparently German website Heise Online has quoted IBM’s Vice President of Deep Computing, David Turek, saying the planned successor to the current processor, slated to have two PowerPC processors and 32 SPEs, will not be released. Extrapolated, then, you can summise that the Cell processor line is terminated, with the current PoweXCell 8i the last one off the production line.</blockquote>
News Source: <a href="http://www.thesixthaxis.com/2009/11/23/cell-processor-dead-in-the-water/" target="_blank">TheSixthAxis</a>
tech3475
11-23-2009, 07:06 AM
I bet Gabe Newell's pleased with himself.
I wonder if they are just going to make the PS4 more dev friendly this time at launch (the cell is powerful but look at all the rubbish ports the first two years of its life).
billysastard
11-23-2009, 07:11 AM
*waves goodbye to any chance backwards compatibility *
Trashcat
11-23-2009, 07:22 AM
It was born dead. How many devices actually used this garbage? And now when nVidia's and ATi's entry level solutions can play betamax2 at the fraction of the cost and power consumption while their top tier solutions provide unchallenged computing power no one will ever need CELL2 and undead CELL1 can crawl back into the grave where it belongs. sony next time put a CPU made for games into your console, and not whatever garbage you have left over that nobody wants to buy from you.
CharmedonWB
11-23-2009, 07:27 AM
haha is all i've got to say especially when i can remember sony claiming the cell chip is the future
Devilish
11-23-2009, 07:28 AM
Pulling out of the development of the Cell processor doesn't mean the end, as Mr. Turek explained, as features of the Cell will continue to be integrated into other processor designs. Without any specific mention of an upcoming model, the company has mentioned that it plans on developing a hybrid technology, which will incorporate both graphics and processing power on a single chip. This is a direction that has been set by NVIDIA, and the launch of its much-anticipated Fermi-based graphics cards.
The news isn't necessarily bad for Sony, which can still hire IBM to create a Cell processor for the company's next-generation console, as this doesn't mean IBM will be involved in the development of Cell, outside the future PlayStation 4 console.
I think this quote is enough to stop the criticism.
Trashcat
11-23-2009, 07:58 AM
I think this quote is enough to stop the criticism.
What criticism? Only a raging sony fanboy who made a CELL tattoo on his forehead back in the day would criticize this, in fact its the first sane decision sony made on console hardware in 10 years.
It was born dead. How many devices actually used this garbage? And now when nVidia's and ATi's entry level solutions can play betamax2 at the fraction of the cost and power consumption while their top tier solutions provide unchallenged computing power no one will ever need CELL2 and undead CELL1 can crawl back into the grave where it belongs. sony next time put a CPU made for games into your console, and not whatever garbage you have left over that nobody wants to buy from you.
You do know that the Cell was designed specifically for the PS3 right? how is that leftovers? It was built from the ground up as a full-custom microprocessor for game physics and AI.
Bet you can't even tell me why any other processor is more suited to games, especially when the Cell, Xenon, and broadway are all PowerPC based at heart. if you think that Xenon or Broadway are going to be used in future consoles I'd like to bet a small sum that they're not. guess that would leave them undead and crawling into a hole in your eyes too.
Trashcat
11-23-2009, 08:46 AM
You do know that the Cell was designed specifically for the PS3 right?
No it wasnt.
It was built from the ground up as a full-custom microprocessor for game physics and AI.
It was build for stream data processing if you know what that is.
Bet you can't even tell me why any other processor is more suited to games, especially when the Cell, Xenon, and broadway are all PowerPC based at heart.
The fact that you even mention 10 year old Broadway tells me a lot. To anyone who knows something about modern CPU architectures and game engines it should be clear which one is better for games just from reading the specs.
if you think that Xenon or Broadway are going to be used in future consoles I'd like to bet a small sum that they're not.
I sure hope none of them will be used any more.
crypysmoker
11-23-2009, 09:24 AM
Funny to see why people think its dead.
You guys do understand it boils down to UTTER LAZINESS.
Developers have been working on normal CPU architecture for years and used to it. Then a new type comes along and noone knows what to do with it.
If more people would have learned how to use it, then all the PS3 would need is about another 512mb vram and it would murder the 360.
The cell processor is a very very good idea.... If developers werent so stuck in their damn ways.
Trashcat
11-23-2009, 09:34 AM
Funny to see why people think its dead.
You guys do understand it boils down to UTTER LAZINESS.
Developers have been working on normal CPU architecture for years and used to it. Then a new type comes along and noone knows what to do with it.
If more people would have learned how to use it, then all the PS3 would need is about another 512mb vram and it would murder the 360.
The cell processor is a very very good idea.... If developers werent so stuck in their damn ways.
Please please say you are just trolling...
No it wasnt.
It was build for stream data processing if you know what that is.
The fact that you even mention 10 year old Broadway tells me a lot. To anyone who knows something about modern CPU architectures and game engines it should be clear which one is better for games just from reading the specs.
I sure hope none of them will be used any more.
http://www.research.ibm.com/cell/
that was one of its goals during its design stage, still haven't told me how it was a leftover? It's also funny since the PPE core they designed during the Cell project is essentially the one they used in the 360 xenon bar a modification. Was the 360 processor dead on arrival too or does that not apply because you can't make a stupid betamax2 reference?
stream data processing? lol yes please enlighten me. do inform me about why this does not apply to gaming exactly, especially physics?
What's even more funny is how an article about long string pattern matching of large working sets (something that does not really apply to a standalone console in any way) somehow implies that the Cell is unfit for gaming consoles in your eyes. Hypocrite.
What does it tell you? yes please do read the specs because you haven't told me anything yet about why one architecture is more suited to gaming than the other, other than ease for the low-level programmer.
Cloudkill
11-23-2009, 09:54 AM
The main issue with the PS3 was the CPU in the first place what a bottle neck lets spend billions on building and CPU then put a 2 year old graphics card in the unit.
It would have been better off and cheaper to develop either going down the skull trail route, or just a single quad core cpu and much better graphics card.
Lets face it you could use the skull trail system and have intel design two quad cores that have different instructions built into the both of them. I can promise the graphics of the system would have been exactly the same as it is today.
PPC CPU's seem to have a curse on them since Apple moved over to intel.
The main issue with the PS3 was the CPU in the first place what a bottle neck lets spend billions on building and CPU then put a 2 year old graphics card in the unit.
It would have been better off and cheaper to develop either going down the skull trail route, or just a single quad core cpu and much better graphics card.
Lets face it you could use the skull trail system and have intel design two quad cores that have different instructions built into the both of them. I can promise the graphics of the system would have been exactly the same as it is today.
PPC CPU's seem to have a curse on them since Apple moved over to intel.
That would have made it twice as expensive for the consumer especially if the 10 year plan is true. Skull trail is already expensive for what you actually get. Not sure what you mean by the skull trail route but if you mean two CPUs I would ask why? that would make it twice as expensive due to your die size alone not to mention the fact that a single chip is faster if you wanted to keep the same overall size.
Please please say you are just trolling...
No I think you're the one doing that.
ThreeDog
11-23-2009, 10:16 AM
If anyone uses something as complicated as CELL in their next machine i won't be buying as early as i did the PS3. I am not calling CELL rubbish i like what it can do but combined with poor decisions and expenses it affected the platform negatively.. all that after developers complained how hard the PS2 was to develop for.
Sony has to do it right next time, it would make sense to round up developers ask them what they want and need i honestly believe this is the true path to success.
MadMax31
11-23-2009, 10:19 AM
Funny to see why people think its dead.
You guys do understand it boils down to UTTER LAZINESS.
Developers have been working on normal CPU architecture for years and used to it. Then a new type comes along and noone knows what to do with it.
If more people would have learned how to use it, then all the PS3 would need is about another 512mb vram and it would murder the 360.
The cell processor is a very very good idea.... If developers werent so stuck in their damn ways.
If you were a developer using new architecture, crappy SDK and little to non support from Sony and all the pressure from the corporates you would b*tch with the lazy developers like the rest of the kids that don't have a clue what "a tight schedule" or "meeting deadlines" or even "profit" means.
The only ones to blame here are the source (Sony), they should have developed better SDK and/or support to 3rd parties. Any hardware is as good as the software that back it's up.
tech3475
11-23-2009, 10:22 AM
The other problem for sony is that there hardware is getting more complicated all the time.
One reason why we are yet to see a software based emulator for the PS2 is because that is difficult to emulate.
Imagine if they tried putting BC in the PS4 without including CELL for BC unless its architecturally compatible.
If anyone uses something as complicated as CELL in their next machine i won't be buying as early as i did the PS3. I am not calling CELL rubbish i like what it can do but combined with poor decisions and expenses it affected the platform negatively.. all that after developers complained how hard the PS2 was to develop for.
Sony has to do it right next time, it would make sense to round up developers ask them what they want and need i honestly believe this is the true path to success.
Surely that's even more of a reason to include the Cell in a new console since the low-level programmers have mastered it. If it were a new architecture it would actually be worse. Not to mention costs will be lower if they used the Cell again since they don't have to pump another $400M in architecture design like they did with the Cell. Cell R&D would have broke even and they wouldn't need to recoup costs either so the actual chip would be cheaper.
QrafTee
11-23-2009, 10:52 AM
So that means that the PS4 will support quantum computing, right? :D
Patango
11-23-2009, 11:10 AM
You do know that the Cell was designed specifically for the PS3 right?
It absolutely was not designed for PS3.
how is that leftovers? It was built from the ground up as a full-custom microprocessor for game physics and AI.
Complete nonsense, it was designed for media stream compression/decompression and was primarily to be used in white goods (TV's, TV boxes, PVR's etc...). PS3 was (at best) an afterthought.
Bet you can't even tell me why any other processor is more suited to games, especially when the Cell, Xenon, and broadway are all PowerPC based at heart.
Because they're cheap, not because they're specifically suited to games.
if you think that Xenon or Broadway are going to be used in future consoles I'd like to bet a small sum that they're not. guess that would leave them undead and crawling into a hole in your eyes too.
No-one ever said it was, except Sony, they stated it would be the end of x86!!
You'd imagine they would have learned their lesson after the "Graphics Synthesizer" on the PS2.
Patango
11-23-2009, 11:24 AM
Surely that's even more of a reason to include the Cell in a new console since the low-level programmers have mastered it.
When "mastering it" means your still spending lots of money and effort to get the same out as you would another chip in 1/10 of the time, you can see why it makes sense to ditch it.
If it were a new architecture it would actually be worse.
Only if that new architecture uses the same "round the houses" approach that Sony have had since the PS2.
Not to mention costs will be lower if they used the Cell again since they don't have to pump another $400M in architecture design like they did with the Cell. Cell R&D would have broke even and they wouldn't need to recoup costs either so the actual chip would be cheaper.
Or you could just buy a CPU off the shelf that is already much much faster than the one you were going to develop in house.
kneehighspy
11-23-2009, 11:50 AM
No it wasnt.
It was build for stream data processing if you know what that is.
The fact that you even mention 10 year old Broadway tells me a lot. To anyone who knows something about modern CPU architectures and game engines it should be clear which one is better for games just from reading the specs.
I sure hope none of them will be used any more.
qft. the cell was originally based upon the power pc core and was originally inteneded for ibm's new server line, but pulled into the ps3 project by the same team that was on the 360s cpu, the teams were then split independently.
the book, the race for a new game machine, explains in great details about the developement of the cell and 360 processors.
It absolutely was not designed for PS3.
Complete nonsense, it was designed for media stream compression/decompression and was primarily to be used in white goods (TV's, TV boxes, PVR's etc...). PS3 was (at best) an afterthought.
I don't understand how a processor which was designed from the ground up and they knew they would use it in the PS3 from the beginning of the project could be considered an afterthought. Just because it's good at compression and decompression so has more than one application does not make its primary design goal white goods it had a number of applications but it's not an afterthought if you thought about using it in the PS3 while designing it.
Because they're cheap, not because they're specifically suited to games.
sure you can make bigger (more expensive) better gaming processors but it seems you are trying to give me irrelevant points rather than look at my point
"sony next time put a CPU made for games into your console"
PS3, 360 and Wii are all PowerPC based and the 360 processor is based on the Cell PPE design. Is he saying that MS, Nintendo and Sony should all ditch PPC in favour of more expensive processors for all our consoles? doubt it.
No-one ever said it was, except Sony, they stated it would be the end of x86!! What fools. You'd imagine they would have learned their lesson after the "Graphics Synthesizer" on the PS2.
Yes but again more irrelevant points, his view was that the Cell is not going to be in future IBM servers (not even confirmed not to be in PS4) and has a successor so that somehow means it was dead on arrival. Bet he doesn't think the same for the xenon or broadway when they probably aren't going to be in future devices either.
But in reply to your point Cell-Like processors (obviously not the Cell itself) were in intels roadmap so yes they are thinking that way but calling people fools for a wrong prediction is nothing but hindsight in the fools favor.
http://www.anandtech.com/printarticle.aspx?i=2379
"We will never make a 32-bit operating system, but I'll always love IBM"
somebody once said that but I can assure you he knows more about the market than anybody here not to mention he has made a lot more money.
When "mastering it" means your still spending lots of money and effort to get the same out as you would another chip in 1/10 of the time, you can see why it makes sense to ditch it.
Actually its peak performance is higher so no not the same results but if you're talking about the overall system blame the GPU used not the CPU.
Only if that new architecture uses the same "round the houses" approach that Sony have had since the PS2.
Or you could just buy a CPU off the shelf that is already much much faster than the one you were going to develop in house. you could do I guess but manufacturing and royalties become your problem instead. So you might think using an expensive non-PPC processor is a good idea but I don't think it is. opinion I guess.
qft. the cell was originally based upon the power pc core and was originally inteneded for ibm's new server line, but pulled into the ps3 project by the same team that was on the 360s cpu, the teams were then split independently.
the book, the race for a new game machine, explains in great details about the developement of the cell and 360 processors.
Haven't read the book but wasn't STI set up for designing the cell from the beginning so Sony and Toshiba were there from the beginning of the Cell project during which IBM were approached by MS. Just saying because I always thought The Cell project team included Sony and toshiba and the application of using it in the PS3 from the very beginning.
When the companies entered into their partnership in 2001, Sony, Toshiba and IBM committed themselves to spending $400 million over five years to design the Cell, not counting the millions of dollars it would take to build two production facilities for making the chip itself. IBM provided the bulk of the manpower, with the design team headquartered at its Austin, Texas, offices. Sony and Toshiba sent teams of engineers to Austin to live and work with their partners in an effort to have the Cell ready for the Playstation 3's target launch, Christmas 2005.
But a funny thing happened along the way: A new "partner" entered the picture. In late 2002, Microsoft approached IBM about making the chip for Microsoft's rival game console, the (as yet unnamed) Xbox 360. In 2003, IBM's Adam Bennett showed Microsoft specs for the still-in-development Cell core. Microsoft was interested and contracted with IBM for their own chip, to be built around the core that IBM was still building with Sony.
All three of the original partners had agreed that IBM would eventually sell the Cell to other clients. But it does not seem to have occurred to Sony that IBM would sell key parts of the Cell before it was complete and to Sony's primary videogame-console competitor. The result was that Sony's R&D money was spent creating a component for Microsoft to use against it.
Mr. Shippy and Ms. Phipps detail the resulting absurdity: IBM employees hiding their work from Sony and Toshiba engineers in the cubicles next to them; the Xbox chip being tested a few floors above the Cell design teams. Mr. Shippy says that he felt "contaminated" as he sat down with the Microsoft engineers, helping them to sketch out their architectural requirements with lessons learned from his earlier work on Playstation.
The deal only got worse for Sony. Both designs were delivered on time to IBM's manufacturing division, but there was a problem with the first chip run. Microsoft had had the foresight to order backup manufacturing capacity from a third party. Sony did not and had to wait another six weeks to get their first chips. So Microsoft actually got the chip that Sony helped design before Sony did. In the end, Microsoft's Xbox 360 hit its target launch in November 2005, becoming its own success. Because of various delays, the Playstation 3 was pushed back a full year
ThreeDog
11-23-2009, 02:10 PM
Surely that's even more of a reason to include the Cell in a new console since the low-level programmers have mastered it. If it were a new architecture it would actually be worse. Not to mention costs will be lower if they used the Cell again since they don't have to pump another $400M in architecture design like they did with the Cell. Cell R&D would have broke even and they wouldn't need to recoup costs either so the actual chip would be cheaper.
For a game of true quality the length of development start to going gold is still too long and there are a few key studios/publishers that are ignoring the PS3 due to the extra work involved, a few have gone off the idea since they tried it stating other development kits are easier to work with and take less time.
Look at it this way, it may not be wise for Sony (debatable) but it is in IBMs best interest. The first yield was said to have been pretty bad combined with the issue of no backup, Sony is the primary purchaser of these chips they got the best deal and most other products didn't get off the ground, IBM paid for most of the shrinks and Sony sold its stake in cell to toshiba when hd-dvd crashed.
It probably cost all parties a fair amount more money than expected but since the other uses for CELL didn't get far why not make something dedicated to gaming that's easier for developers to deal with. Why spend more money on R&D enhancing CELL when more simplistic solutions are said to be on their way?
I'm just going to throw it out there maybe it just wouldn't work but why not Fermi or maybe a way off for what's required something along the lines of Larrabee?
crypysmoker
11-23-2009, 02:39 PM
Please please say you are just trolling...
No I was sserious. I wouldnt want to take your full time job from you... ;)
For a game of true quality the length of development start to going gold is still too long and there are a few key studios/publishers that are ignoring the PS3 due to the extra work involved, a few have gone off the idea since they tried it stating other development kits are easier to work with and take less time.
Look at it this way, it may not be wise for Sony (debatable) but it is in IBMs best interest. The first yield was said to have been pretty bad combined with the issue of no backup, Sony is the primary purchaser of these chips they got the best deal and most other products didn't get off the ground, IBM paid for most of the shrinks and Sony sold its stake in cell to toshiba when hd-dvd crashed.
It probably cost all parties a fair amount more money than expected but since the other uses for CELL didn't get far why not make something dedicated to gaming that's easier for developers to deal with. Why spend more money on R&D enhancing CELL when more simplistic solutions are said to be on their way?
I'm just going to throw it out there maybe it just wouldn't work but why not Fermi or maybe a way off for what's required something along the lines of Larrabee?
Fermi is still pretty expensive perhaps in the future it won't be but right now judging by their $2,500 price tag for their tesla cards I would say that's not viable even if it were somehow geared towards gaming. sony would have to pay NVIDIA a large sum and that would mean you would have to pay sony a large sum too, and for what exactly? there would be very little difference in difficulty in programming, both would require the programmer to write highly parallel code for specialized cores. To me the simplest option would be to use more of the same rather than starting from scratch, not saying that's what they're going to do but to me that would require the least amount of effort and the least amount of investment.
MadonnaProject
11-23-2009, 05:22 PM
Developers everywhere will heave a sigh of relief. Scattered processing was never a really good idea to begin with. God riddance.
Why go for something different when the normal powerpc standard has worked so well for almost 30 years?
msanchez
11-23-2009, 05:52 PM
Well in all honesty as great as the cell could be, unless developers feel the need to learn it then the cell will never actually "BE" it will always be a "COULD BE". What would sony have to do to make the cell more popular? I think they would have to release the ps4 before the others release their consoles, and do so at a price range that even Trashcat will pre-order one... now the problem with that is what will all those people that have bought ps3s have to say about sony decreasing support for the ps3 so quickly? anybody that's sane understands that the 10 year life thing while true doesn't mean that sony won't release another console 7 - 8 years after the ps3, and then slowly kill off the ps3 as it's doing with the ps2, but if they do that too quickly (which they have to do if they want enough sales to make devs feel they need to learn to write for the cell) then a lot of customers will be pissed at them.
Sadly, sony just needs to cater to the cry babies at least for now... if they ever regain control as they did with the ps2 and ps1 they can try to enforce their hardware again (as long as they don't go and do ps3 launch part 2), but for now sony should just play along. "Live to fight another day" if you will.
EL_CAD
11-23-2009, 06:22 PM
But what about all the cell processors around my house that were supposed to make the games better? Where's the toasters and coffee makers that are going to help render GT5? I guess that plan went the way of being able to output to two HDTVs at once.
But what about all the cell processors around my house that were supposed to make the games better? Where's the toasters and coffee makers that are going to help render GT5? I guess that plan went the way of being able to output to two HDTVs at once.
lol, I keep hearing this toaster comment but where does it originate from.
grapeape
11-23-2009, 07:14 PM
lol, I keep hearing this toaster comment but where does it originate from.
Back when IBM and Sony unveiled the Cell, the keynote mentioned something about a future of computer processing where cell based TV's and toasters would be able to help your computer crunch numbers when describing the coprocessing capability of cell based devices...anyway thats paraphrased but the speech made it sound like the cell was going to take over your house. :) Dont know if thats really where the PS3 toaster stuff happened...but I remember folks poking fun at it on slashdot and comparing it to Skynet. :)
Back when IBM and Sony unveiled the Cell, the keynote mentioned something about a future of computer processing where cell based TV's and toasters would be able to help your computer crunch numbers when describing the coprocessing capability of cell based devices...anyway thats paraphrased but the speech made it sound like the cell was going to take over your house. :) Dont know if thats really where the PS3 toaster stuff happened...but I remember folks poking fun at it on slashdot and comparing it to Skynet. :)
Thanks grape that's another meme I can tick off the checklist.
I saw linked videos to youtube about a ps3 and a toaster that had me thinking "wtf?"
http://www.youtube.com/watch?v=MCj0BSd8grs
I remember during the early years of bluetooth they were saying the same thing. how we would set the oven, toaster, coffee machine or toothbrushes with other bluetooth devices. several years down the line and most TVs are still using IR yet alone getting any sort of remote oven or toaster standard.
kneehighspy
11-25-2009, 12:42 PM
Haven't read the book but wasn't STI set up for designing the cell from the beginning so Sony and Toshiba were there from the beginning of the Cell project during which IBM were approached by MS. Just saying because I always thought The Cell project team included Sony and toshiba and the application of using it in the PS3 from the very beginning.
that's part of the story, ibm was developing the cell idea for the servers when sony came along (before microsoft) and weighed different options on different architectures and liked the cell idea (that was going into future servers) best. so the joint group started on the design of the cell processor to integrate into the ps3. initially sony was going a with their own graphics gpu design, but things kinda faltered and didn't work well with the cell so they outsourced the gpu work to nvidia (using existing 78xx architecture).
then about a year into developement of the cell, microsoft came along wanting a new processor, were shown several ideas (even the cell) and went with the basic powerpc 3 core processor. several key devs had to go back and forth between the different design teams and felt like traders to sony and their project.
it's a great read if you get the chance.
Powered by vBulletin® Version 4.1.5 Copyright © 2013 vBulletin Solutions, Inc. All rights reserved.