Well, the topic of this thread is will the new PowerMacs use Cell Processor. I was pessimistic at first but the more I read Tiger documentation the more I feel that the answer is YES. If we have a broader look at the "political" situation it it looks even more feasible.
Earlier in this thread Programmer noted that programmers need some time to adopt the Cell but I think that the vast majority of them will never bother. There are only relatively small groups of them which will be affected. Those are the guis who are making:
1. Compilers. I think, however, that it is easier to optimize for the Cell than to implement automatic vectorization (well, the Cell will benefit from the later as well...).
2. Operating systems. A decade ago statistics for Mac OS (7?) showed that the average application spends 60% of the time running system functions. For Cocoa applications this should be higher so any OS optimization benefits ALL applications, even those compiled for older system. The new apps which use the Core Data, Core Image, Core ... stuff will benefit the most.
3. Database engines. There are more legacy approaches here that could be easily optimized for Cell
4. Games
5. Video, audio processing modules (not necessarily full applications; the current implementation of Core Image has enough stuff to make a Photoshop Lite in a couple of months, RAW import, ColorSync and CMYK support included).
6. Few other areas - scientific stuff etc.
I have an inside info that IBM plans to start working on 1and 3 and Sony on 4 and 5 some time in the second half of next year.
Mac OS and Linux will cover the 2. I feel that Tiger was build with the Cell in mind.
I think that besides the technical stuff Sony, IBM and Toshiba made some effort beside the scene to ensure broader support for the Cell when it arrives. And I can not imagine the whole picture without Apple in it...
May be it's my paranoia but:
- Steve invited Sony CEO on stage
- Steve says in an interview that major computer manufacturers approached Apple for licensing Mac OS X. Everybody talked about HP and Dell but aren't Sony and Toshiba making notebooks?
\t- IBM sells it's PC business, initiates Power.org
There was also a rumor published by some of the relatively reliable rumor sites (cloud not find the link, could somebody help?) that two independent sources from ATI and nVidia claimed that their companies have approached Apple offering a deal for new high-end graphic cards but Apple was not interested.
I think that IBM, Sony and Toshiba already negotiated their areas of interest. To achieve their goals they need to move the Cell to the enterprise (IBM) and to the consumer products (Sony, Toshiba) and they need to make this FAST. Otherwise the momentum will be lost. They witnessed the Itanic progress, didn't they? Itanic was supposed to be the new Intel processor architecture for a wide range of computers, not a high end a'la Power4/Power5 chip, wasn't it? And how many billions Intel lost on that? Will STI hope that thousands of developers will embrace the Cell rather that play the "wait and see" game or it will take care that some powerful players (preferably those known for their innovation and influence on the industry) will be committed to Cell beforehand? They have the whole history to learn from: Intel thought they are so powerful that it is a no-brainer that they will be successful with BETTER architecture. Sony thought that they are such a big name in entreatment industry they will eat all those apples, 'pods and mp3s for breakfast. We (and most notably, THEY) know what happened...
And here comes Apple. Apple has a lot of technologies which are already in place. If STI starts building them from scratch it will be a great opportunity for the competition to catch up.
So, will new PowerMacs use Cell Processor? - Yes. May be the next year at the earliest, but the work is already in progress...
I think shadow might be right on this one (good pulling together all of the rumors).
Steve's comment about this being the year of HD is true to a point. This year Tiger will debut with H.264 in QT and it will all be about HD.
To truly grab marketshare Gaming though deplorable to Steve is a driving force. All of those kids out there tell their parents what computer they want. Parents buy them. Apple missed out on the gaming market years ago and as such lost most of its market share.
Now, with HD resolution what kinds of games (and other graphic content) can be produced? The game manufacturers will write for the platform that can take advantage of their talents. Pixar will render and produce content on macs (because steve wants it that way) and everyone will be blown away.
We don't know that STI won't include apple as a partner or shareholder after the debut of tiger. Think about apple's marketing.
big bold letters - Tiger with the power of CELL! or some such thing.
Hell apple has made a music player without a screen and gotten the world to look at random songs in their library as wonderful.
Cell in the powermacs first, Perhaps even announced at wwdc this june. Gives the developers something to really go crazy for!
. . . Cell in the powermacs first, Perhaps even announced at wwdc this june. Gives the developers something to really go crazy for! . . .
Ron Okamoto, Apple vice president of Worldwide Developer Relations, told MacCentral, "One of the keys for this year's conference is going to be innovation."
1. Compilers. I think, however, that it is easier to optimize for the Cell than to implement automatic vectorization (well, the Cell will benefit from the later[sic] as well...).
Cell is essentially one PPC with AltiVec and 8 independent SIMD units. IBM has poured a great deal of effort into autovectorization, and they say they'll have it for the next release of their compiler. A similar effort has been ongoing with GCC. The autovectorization code is actually the most plausible adaptation to the Cell architecture, because Cell is all about parallelism (both data level and thread level), and vectors are basically data in parallel.
Quote:
4. Games
With the ever-present caveat that those games which are essentially self-contained and designed to run on a PC (i.e., a single, monolithic, scalar core) are not going to benefit without a lot of clever footwork from the porting houses (and cooperation from the publishers, who might not want the code base forked more than is absolutely necessary). This will get somewhat better when Intel's dual-core chip ships in the latter half of this year and game designers have to confront threading as a real performance enhancement.
Quote:
I have an inside info that IBM plans to start working on 1and 3 and Sony on 4 and 5 some time in the second half of next year.
Inside what? IBM has been making noises about autovectorization in their own compiler, and in gcc, since AltiVec appeared in the 970.
Quote:
To achieve their goals they need to move the Cell to the enterprise (IBM) and to the consumer products (Sony, Toshiba) and they need to make this FAST. Otherwise the momentum will be lost. They witnessed the Itanic progress, didn't they?
Itanic didn't have any momentum because it was bloated, years behind schedule, hilariously underperforming in its first incarnation, and far too power-hungry and expensive in latter incarnations. There was nowhere for it to move except the very high end. If Cell is on time, lives up to expectations performance-wise (including performance/watt), and carries a reasonable price, then momentum will not be a problem. Not with Sony already on board for the PS3.
I'm definitely interested to see what Apple does with Cell, and when, not least because it screams "multimedia."
. . . There was also a rumor published by some of the relatively reliable rumor sites (cloud not find the link, could somebody help?) that two independent sources from ATI and nVidia claimed that their companies have approached Apple offering a deal for new high-end graphic cards but Apple was not interested. . .
If true, this is a strong indicator that Cell is on the way. With Cell to do core services, Apple does not need high-end graphic cards. In Cell based hardware, the graphic card will play a much smaller role than it does now. This is what I suspect from all I read. If I'm wrong, our good experts will hopefully set the record straight.
Cell is essentially one PPC with AltiVec and 8 independent SIMD units. IBM has poured a great deal of effort into autovectorization, and they say they'll have it for the next release of their compiler. A similar effort has been ongoing with GCC. The autovectorization code is actually the most plausible adaptation to the Cell architecture, because Cell is all about parallelism (both data level and thread level), and vectors are basically data in parallel.
That's my poit! They are working on this for years!
Quote:
With the ever-present caveat that those games which are essentially self-contained and designed to run on a PC ...
Isn't Sony supposed to use the chip for gaaming?
Quote:
I'm definitely interested to see what Apple does with Cell, and when, not least because it screams "multimedia."
Apple invested a lot in AltiVec optimization. The Core Data\\Image\\Video... collection suits data parallelism very well. And (it seems) they are moving all their multimedia software and iApps to those frameworks. That makes me think that they have the Cell in mind.
Some of you seem to have been caught up in our own dreaming scheme in here. Cell is not going to be in the NEXT PowerMac no matter how much you bullshit yourself. Cell may wind up in a Future PowerMac, but chances are it wont be in 2005. Thinking anything else is far too optimistic.
snoopy, SPE's are not going to replace a GPU. It could conceivably help, but it hardly seems as wise as to use it for something else, and use Core Image to utilize a real GPU that has been designed by the best graphics engineers to do just that.
Some of you seem to have been caught up in our own dreaming scheme in here. Cell is not going to be in the NEXT PowerMac no matter how much you bullshit yourself. Cell may wind up in a Future PowerMac, but chances are it wont be in 2005. Thinking anything else is far too optimistic.
See my first post earlier on this thread - 2006 at the earliest. I think we are at least 2 generations away fron the CellMac (if it ever arrives).
I am not expert on graphics but I think that Cell may take on the GPU.
link below helps dispell some of the misconceptions about Cell. It doesn't sound like Cell is really as applicable to the PC as we thought or at least according to this guy.
PS3 games are going to be written directly for the cell. PC games ported to the Mac will require a lot of new code to take advantage of the secondary processors. I don't know if it'd be worth the effort to the porting houses.
See my first post earlier on this thread - 2006 at the earliest. I think we are at least 2 generations away fron the CellMac (if it ever arrives).
I am not expert on graphics but I think that Cell may take on the GPU.
I look at it this way. If a GPU is lagging, or has a bottleneck that slows it down. It could be possible to offload some elements that would be caught in said bottleneck to an SPE, (possibly even through a software upodate) but no SPE is going to be as precise, or functional as an entire Nvidia designed GPU. An SPE is how big? How in the world can it calculate all that. A cell based GPU on the other hand could be a possibility in the future.
link below helps dispell some of the misconceptions about Cell. It doesn't sound like Cell is really as applicable to the PC as we thought or at least according to this guy.
P.S. I don't know that much about computers so be please be easy on me.
Finally! This is what I've been saying, but people keep jumping to conclusions with their speculative hypotheses, and I'm in (Fill in the Blank) field of work so I know what I'm saying.
Quote:
What the Cell Isn't
No, it's not your next PC or server.
Information on the new cell processor from Sony and IBM has been somewhat available to the geekily inclined public for quite a while now. So, I was very surprised by the questions I was being asked after the Cell's more official debut at the ISSCC conference. And, I was even more surprised by some of the reports that came out of the conference. Reporters have been consistently trying to make the Cell a competitor to the venerable X86 architecture and the Pentium in particular.
For example, here is the sort of quote you might encounter:
"Analysts say that the difference [between the Pentium and the Cell] is likely to shake up the contemporary model of computing that Microsoft and Intel have fostered. [because] The new chip also has been designed to handle multiple operating systems and programs simultaneously."
The Cell is not a Pentium killer or a competitor?it's a SIMD supercomputer, and just because it has an IBM Power processor in it, which has been demonstrated to run X86 code in an emulation mode, that's not the use or the future of this Cell.
Here's another one:
"Later this year, Intel and Advanced Micro Devices each plans to launch their own 'multicore' chips, which similarly up the number of commands that can be executed at once. IBM and Sun Microsystems already sell such multicore microprocessors, largely for business servers."
The Cell has an array of SIMD RISC processors, not multi-cell X86 processors. Nor will it be a super server chip?it's just not built that way. It's built to process the same kind of data, which is highly predictable, not changing data like a GP CPU in desktop or a server deals with. Mix up the data stream and types and feed it to a Cell and you'll see it screech to a slow crawl. It costs 18 cycles any time a branch mis-predict? that's not server-quality processing.
Is it your next workstation?
IBM Chairman, President, and Chief Executive Samuel Palmisano said that this year it'll sell a workstation containing Cell. That's true, it will be built at IBM's Boeblingen facility near Stuttgart (see Tech Watch, November 29). Such a system will be used for software development for the future Sony console. It's possible it may be used for scientific work, but special compilers and applications will have to be created to make use of it. That comes under the "someday" heading.
Is it a superomputer, on a chip?
Well yes, architecturally, it's exactly like the multi-processor SIMD vector machines that were built in the mid-eighties, and to some extent are still being built (only 4.2% of today's supercomputers are vector based). However, at 256 GFLOPS it's not, by itself, in the supercomputer class. (The slowest machine on the current list of the Top 500 supercomputers does 624 GFLOPS http://www.top500.org/lists/plists.p...5&M=06&Y=2004.) But you can envision an array of Cells being made into a supercomputer, and given the expected low cost of this soon-to-be-mass-produced chip, that's entirely feasible, except that it might wrinkle the feathers of IBM's BlueGene group a bit.
It's not a Grid computer
The other popular misconception, which IBM and Sony are doing nothing to dispel, is the notion that the Cell will hook up to some type of grid (like Butterfly net) and either tap into more resources or allow millions of user to simultaneously play a game. Yes, online gaming can take advantage of the Butterfly net, and yes, IBM and Sony said they would use it. But it's not to share Cell resources?it's to play against others just like any other online network.
Take a look at this quote from a story published by CNET:
"Cells can even roam over a network, allowing the processor to perform a type of distributed or grid computing, an increasingly popular enterprise technique in which demanding tasks are divvied up among a gang of networked computers."
Clearly, there are some major misunderstandings about what networked and Grid computing are all about. Grid is used for collaboration and uses the Internet, not the fastest data exchange network in the world and certainly not fast enough for distributed for real-time computer, the type of computing that epitomizes a game.
A Playstation 3 could borrow unused processing power from other consoles on a network, for example, to complete a demanding task such as delivering streaming video. Sony has done a masterful job at creating a science fiction scenario that ignorant journalists have latched onto. The only problem is some of those ignorant journalists have good memories, and in a year or two when the next-generation Playstation is delivered and it doesn't do all these suggested whiz-bang things, those same journalists are going to feel used, and then they will write searing stories about how Sony and IBM have failed to live up to the expectations and promise of the lauded Cell. This misdirected PR BS is going to haunt IBM and Sony. Lest we forget, when Sony introduced the Playstation 2 Emotion Engine it was promised to power a wide variety of consumer electronics, and so far it's failed to live up to that promise.
Is it your next TV?
Well, Toshiba has been quoted as saying it plans to incorporate it into high-end TVs. Given the crashing ASPs on TVs these days, it's not likely the Cell, in its current configuration, will show up inside a TV, high-end or otherwise. The vector processors could be useful for MPEG decoding/encoding, but you wouldn't need eight of them, so a stripped-down version, with maybe two SPEs and a Power CPU, might be made for high-end TVs sometime in 2006, maybe.
So what is a Cell?
It is an eight-IMD processor array with a powerful master controller and program execution GP CPU. Each SIMD has local memory, high-speed interconnects, and a segmented memory structure ideally suited for graphics and codec applications. It has a super-high-speed chip memory interface and an equally fast flexible I/O for the graphics processor. It is currently running Linux and could run other OS's on the GP CPU.
It will be a really great game console processor and general multimedia accelerator, and should cost out at a reasonable consumer price range.?Jon Peddie
link below helps dispell some of the misconceptions about Cell. It doesn't sound like Cell is really as applicable to the PC as we thought or at least according to this guy.
P.S. I don't know that much about computers so be please be easy on me.
Ha! This guy (follow the link) says:
The Cell is not a Pentium killer or a competitor?it's a SIMD supercomputer, and just because it has an IBM Power processor in it, which has been demonstrated to run X86 code in an emulation mode, that's not the use or the future of this Cell.
...
The Cell has an array of SIMD RISC processors, not multi-cell X86 processors.
He understands Intel competion = Pentium emiulation
As BenRoethig pointed out, I wasn't referring to Cell's theoretical capability. Obviously, anything going in a PlayStation has to be suitable for games.
But nearly all the A-list games for the Mac are ported over from Windows, which means they're designed and implemented to run well on Windows, which means they're single-threaded applications optimized to run on one powerful, scalar core, which means they'll run poorly on Cell (especially relative to its theoretical capability). Any optimization for Cell would be up to the porting houses, and it might be impractical or even impossible given the codebase, and time and money constraints, and other issues.
Therefor, practically speaking, Cell would not bring much to the Mac as far as mainstream PC gaming is concerned.
I wouldn't hope for ports from the PS3, either. A Mac game does well if it sells in the tens of thousands. For most console games, that's margin-of-error range, not worth the trouble and cost. In fact, if the threads I've read are any indication, you'd actually target more Mac gamers with the console version...
Quote:
Apple invested a lot in AltiVec optimization. The Core Data\\Image\\Video... collection suits data parallelism very well. And (it seems) they are moving all their multimedia software and iApps to those frameworks. That makes me think that they have the Cell in mind.
Possibly, but not necessarily. After all, all that stuff runs really well on the GPU. Apple can afford to wait and see, or have IBM make them a custom configuration.
Okay, I'm way behind in this thread (as well as my work) and will try and catch up as time permits, but one thing that hits out to me in this article is this quote,
Quote:
The Core Image technology due to appear in OS X "Tiger" already uses GPUs (Graphics Processor Units) for things other than 3D computations and this same technology could be retargeted at the Cell's APUs. Perhaps that's why it was there in the first place...
The barrier to CELL discussed earlier in this thread and which is the same old problem going forward with new tech, is the need to re-write apps, modify the OS, etc, all of which can be a painful, arduous and expensive process.
Could Core Audio, Core Image, and Core Data already be ready to be offloaded to different cells. I'm just a pixel pusher, but it seems to me that this would be feasible with current apps working fine with the core PPC 970 + Altivec and then the developer implementing over time Core Audio, Core Image, and Core Data API's without much of a hiccup?
Could this be part of Apple's plan all along or am I way of base?
Sorry, if someone down stream has already commented about this.
Okay, I'm way behind in this thread (as well as my work) and will try and catch up as time permits, but one thing that hits out to me in this article is this quote,
The barrier to CELL discussed earlier in this thread and which is the same old problem going forward with new tech, is the need to re-write apps, modify the OS, etc, all of which can be a painful, arduous and expensive process.
Could Core Audio, Core Image, and Core Data already be ready to be offloaded to different cells. I'm just a pixel pusher, but it seems to me that this would be feasible with current apps working fine with the core PPC 970 + Altivec and then the developer implementing over time Core Audio, Core Image, and Core Data API's without much of a hiccup?
Could this be part of Apple's plan all along or am I way of base?
Sorry, if someone down stream has already commented about this.
To a pixel pusher. Apples plan was, and is to offload 2D, and 3D graphics work to the GPU using core image API. It's all been stated at WWDC 2004 (video still available at Apple.com/quicktime), and in developer documentation. Using any new tech like this usually extremely simple being that it's probably part of a system -.framework which for the most part is like drag, and drop mach O, and OO programming.
That was my interpretation of Apples presentation from WWDC, but feel free to correct any mistakes.
To a pixel pusher. Apples plan was, and is to offload 2D, and 3D graphics work to the GPU using core image API. It's all been stated at WWDC 2004 (video still available at Apple.com/quicktime), and in developer documentation. Using any new tech like this usually extremely simple being that it's probably part of a system -.framework which for the most part is like drag, and drop mach O, and OO programming.
That was my interpretation of Apples presentation from WWDC, but feel free to correct any mistakes.
I am well aware of Apples plans and I understand the intent to offload 2D and 3D graphics to the GPU using the Core Image API, however my point was based off the comment by the author of the article where he speculated that
Quote:
this same technology could be retargeted at the Cell's APUs.
The point being that to utilize CELL technology would be infact relatively painless for the same reason you point out.
Quote:
Using any new tech like this usually extremely simple being that it's probably part of a system -.framework which for the most part is like drag, and drop mach O, and OO programming.
One reason not to use CELL often assumed would be the huge amount of effort it would take to rewrite everything. The authors point that the Core frameworks could be "retargeted" to use CELL instead of the GPU is what I found interesting.
As I mentioned before, I'm just a pixel pusher and what I thought was an interesting read was pointed out by Programer to be by an author that was full of sh. If that's actually the case or if Programer woke up on the wrong side of the desk, I don't know.
One reason not to use CELL often assumed would be the huge amount of effort it would take to rewrite everything. The authors point that the Core frameworks could be "retargeted" to use CELL instead of the GPU is what I found interesting.
Well yes the author is a total dreamer if he thinks it's going to happen suddenly, or maybe he is speaking of a future cell processor. But, that would be the I/O's for the GPU to presumably offload operations he is referring to. I don't think it's necessary. A good nVidia GPU wont need it, but lesser GPU could conceivably use it, but not this one any time soon.
This particular cell is not going into a Mac ever. When, and if one eventually did I think by the time it does it will be completely different, and evolved from what it is now. - Which is basically a diagram/structure to base the PS3 on. The author is stirring a muck of assumption, presumption, and not interjecting the most critical points to his hypothesis, or he is just taking bit's, and pieces of what everyone else says, and spitting them all out at once like they all belong together, and make sense now without any of the foundation material. But if I say much more about it I'll be doing the same thing because I have not seen his article, only what you copied.
Comments
Earlier in this thread Programmer noted that programmers need some time to adopt the Cell but I think that the vast majority of them will never bother. There are only relatively small groups of them which will be affected. Those are the guis who are making:
1. Compilers. I think, however, that it is easier to optimize for the Cell than to implement automatic vectorization (well, the Cell will benefit from the later as well...).
2. Operating systems. A decade ago statistics for Mac OS (7?) showed that the average application spends 60% of the time running system functions. For Cocoa applications this should be higher so any OS optimization benefits ALL applications, even those compiled for older system. The new apps which use the Core Data, Core Image, Core ... stuff will benefit the most.
3. Database engines. There are more legacy approaches here that could be easily optimized for Cell
4. Games
5. Video, audio processing modules (not necessarily full applications; the current implementation of Core Image has enough stuff to make a Photoshop Lite in a couple of months, RAW import, ColorSync and CMYK support included).
6. Few other areas - scientific stuff etc.
I have an inside info that IBM plans to start working on 1and 3 and Sony on 4 and 5 some time in the second half of next year.
Mac OS and Linux will cover the 2. I feel that Tiger was build with the Cell in mind.
I think that besides the technical stuff Sony, IBM and Toshiba made some effort beside the scene to ensure broader support for the Cell when it arrives. And I can not imagine the whole picture without Apple in it...
May be it's my paranoia but:
- Steve invited Sony CEO on stage
- Steve says in an interview that major computer manufacturers approached Apple for licensing Mac OS X. Everybody talked about HP and Dell but aren't Sony and Toshiba making notebooks?
\t- IBM sells it's PC business, initiates Power.org
There was also a rumor published by some of the relatively reliable rumor sites (cloud not find the link, could somebody help?) that two independent sources from ATI and nVidia claimed that their companies have approached Apple offering a deal for new high-end graphic cards but Apple was not interested.
I think that IBM, Sony and Toshiba already negotiated their areas of interest. To achieve their goals they need to move the Cell to the enterprise (IBM) and to the consumer products (Sony, Toshiba) and they need to make this FAST. Otherwise the momentum will be lost. They witnessed the Itanic progress, didn't they? Itanic was supposed to be the new Intel processor architecture for a wide range of computers, not a high end a'la Power4/Power5 chip, wasn't it? And how many billions Intel lost on that? Will STI hope that thousands of developers will embrace the Cell rather that play the "wait and see" game or it will take care that some powerful players (preferably those known for their innovation and influence on the industry) will be committed to Cell beforehand? They have the whole history to learn from: Intel thought they are so powerful that it is a no-brainer that they will be successful with BETTER architecture. Sony thought that they are such a big name in entreatment industry they will eat all those apples, 'pods and mp3s for breakfast. We (and most notably, THEY) know what happened...
And here comes Apple. Apple has a lot of technologies which are already in place. If STI starts building them from scratch it will be a great opportunity for the competition to catch up.
So, will new PowerMacs use Cell Processor? - Yes. May be the next year at the earliest, but the work is already in progress...
Steve's comment about this being the year of HD is true to a point. This year Tiger will debut with H.264 in QT and it will all be about HD.
To truly grab marketshare Gaming though deplorable to Steve is a driving force. All of those kids out there tell their parents what computer they want. Parents buy them. Apple missed out on the gaming market years ago and as such lost most of its market share.
Now, with HD resolution what kinds of games (and other graphic content) can be produced? The game manufacturers will write for the platform that can take advantage of their talents. Pixar will render and produce content on macs (because steve wants it that way) and everyone will be blown away.
We don't know that STI won't include apple as a partner or shareholder after the debut of tiger. Think about apple's marketing.
big bold letters - Tiger with the power of CELL! or some such thing.
Hell apple has made a music player without a screen and gotten the world to look at random songs in their library as wonderful.
Cell in the powermacs first, Perhaps even announced at wwdc this june. Gives the developers something to really go crazy for!
my
.o2
Originally posted by TednDi
. . . Cell in the powermacs first, Perhaps even announced at wwdc this june. Gives the developers something to really go crazy for! . . .
Ron Okamoto, Apple vice president of Worldwide Developer Relations, told MacCentral, "One of the keys for this year's conference is going to be innovation."
Originally posted by shadow
1. Compilers. I think, however, that it is easier to optimize for the Cell than to implement automatic vectorization (well, the Cell will benefit from the later[sic] as well...).
Cell is essentially one PPC with AltiVec and 8 independent SIMD units. IBM has poured a great deal of effort into autovectorization, and they say they'll have it for the next release of their compiler. A similar effort has been ongoing with GCC. The autovectorization code is actually the most plausible adaptation to the Cell architecture, because Cell is all about parallelism (both data level and thread level), and vectors are basically data in parallel.
4. Games
With the ever-present caveat that those games which are essentially self-contained and designed to run on a PC (i.e., a single, monolithic, scalar core) are not going to benefit without a lot of clever footwork from the porting houses (and cooperation from the publishers, who might not want the code base forked more than is absolutely necessary). This will get somewhat better when Intel's dual-core chip ships in the latter half of this year and game designers have to confront threading as a real performance enhancement.
I have an inside info that IBM plans to start working on 1and 3 and Sony on 4 and 5 some time in the second half of next year.
Inside what? IBM has been making noises about autovectorization in their own compiler, and in gcc, since AltiVec appeared in the 970.
To achieve their goals they need to move the Cell to the enterprise (IBM) and to the consumer products (Sony, Toshiba) and they need to make this FAST. Otherwise the momentum will be lost. They witnessed the Itanic progress, didn't they?
Itanic didn't have any momentum because it was bloated, years behind schedule, hilariously underperforming in its first incarnation, and far too power-hungry and expensive in latter incarnations. There was nowhere for it to move except the very high end. If Cell is on time, lives up to expectations performance-wise (including performance/watt), and carries a reasonable price, then momentum will not be a problem. Not with Sony already on board for the PS3.
I'm definitely interested to see what Apple does with Cell, and when, not least because it screams "multimedia."
Originally posted by shadow
. . . There was also a rumor published by some of the relatively reliable rumor sites (cloud not find the link, could somebody help?) that two independent sources from ATI and nVidia claimed that their companies have approached Apple offering a deal for new high-end graphic cards but Apple was not interested. . .
If true, this is a strong indicator that Cell is on the way. With Cell to do core services, Apple does not need high-end graphic cards. In Cell based hardware, the graphic card will play a much smaller role than it does now. This is what I suspect from all I read. If I'm wrong, our good experts will hopefully set the record straight.
[i]Originally posted by Amorph
Cell is essentially one PPC with AltiVec and 8 independent SIMD units. IBM has poured a great deal of effort into autovectorization, and they say they'll have it for the next release of their compiler. A similar effort has been ongoing with GCC. The autovectorization code is actually the most plausible adaptation to the Cell architecture, because Cell is all about parallelism (both data level and thread level), and vectors are basically data in parallel.
That's my poit! They are working on this for years!
With the ever-present caveat that those games which are essentially self-contained and designed to run on a PC ...
Isn't Sony supposed to use the chip for gaaming?
I'm definitely interested to see what Apple does with Cell, and when, not least because it screams "multimedia."
Apple invested a lot in AltiVec optimization. The Core Data\\Image\\Video... collection suits data parallelism very well. And (it seems) they are moving all their multimedia software and iApps to those frameworks. That makes me think that they have the Cell in mind.
snoopy, SPE's are not going to replace a GPU. It could conceivably help, but it hardly seems as wise as to use it for something else, and use Core Image to utilize a real GPU that has been designed by the best graphics engineers to do just that.
High end graphics cards will still be needed.
Originally posted by onlooker
Some of you seem to have been caught up in our own dreaming scheme in here. Cell is not going to be in the NEXT PowerMac no matter how much you bullshit yourself. Cell may wind up in a Future PowerMac, but chances are it wont be in 2005. Thinking anything else is far too optimistic.
See my first post earlier on this thread - 2006 at the earliest. I think we are at least 2 generations away fron the CellMac (if it ever arrives).
I am not expert on graphics but I think that Cell may take on the GPU.
http://www.jonpeddie.com/index.shtml
P.S. I don't know that much about computers so be please be easy on me.
Originally posted by shadow
Isn't Sony supposed to use the chip for gaming?
PS3 games are going to be written directly for the cell. PC games ported to the Mac will require a lot of new code to take advantage of the secondary processors. I don't know if it'd be worth the effort to the porting houses.
Originally posted by shadow
See my first post earlier on this thread - 2006 at the earliest. I think we are at least 2 generations away fron the CellMac (if it ever arrives).
I am not expert on graphics but I think that Cell may take on the GPU.
I look at it this way. If a GPU is lagging, or has a bottleneck that slows it down. It could be possible to offload some elements that would be caught in said bottleneck to an SPE, (possibly even through a software upodate) but no SPE is going to be as precise, or functional as an entire Nvidia designed GPU. An SPE is how big? How in the world can it calculate all that. A cell based GPU on the other hand could be a possibility in the future.
Originally posted by cben
link below helps dispell some of the misconceptions about Cell. It doesn't sound like Cell is really as applicable to the PC as we thought or at least according to this guy.
http://www.jonpeddie.com/index.shtml
P.S. I don't know that much about computers so be please be easy on me.
Finally! This is what I've been saying, but people keep jumping to conclusions with their speculative hypotheses, and I'm in (Fill in the Blank) field of work so I know what I'm saying.
What the Cell Isn't
No, it's not your next PC or server.
Information on the new cell processor from Sony and IBM has been somewhat available to the geekily inclined public for quite a while now. So, I was very surprised by the questions I was being asked after the Cell's more official debut at the ISSCC conference. And, I was even more surprised by some of the reports that came out of the conference. Reporters have been consistently trying to make the Cell a competitor to the venerable X86 architecture and the Pentium in particular.
For example, here is the sort of quote you might encounter:
"Analysts say that the difference [between the Pentium and the Cell] is likely to shake up the contemporary model of computing that Microsoft and Intel have fostered. [because] The new chip also has been designed to handle multiple operating systems and programs simultaneously."
The Cell is not a Pentium killer or a competitor?it's a SIMD supercomputer, and just because it has an IBM Power processor in it, which has been demonstrated to run X86 code in an emulation mode, that's not the use or the future of this Cell.
Here's another one:
"Later this year, Intel and Advanced Micro Devices each plans to launch their own 'multicore' chips, which similarly up the number of commands that can be executed at once. IBM and Sun Microsystems already sell such multicore microprocessors, largely for business servers."
The Cell has an array of SIMD RISC processors, not multi-cell X86 processors. Nor will it be a super server chip?it's just not built that way. It's built to process the same kind of data, which is highly predictable, not changing data like a GP CPU in desktop or a server deals with. Mix up the data stream and types and feed it to a Cell and you'll see it screech to a slow crawl. It costs 18 cycles any time a branch mis-predict? that's not server-quality processing.
Is it your next workstation?
IBM Chairman, President, and Chief Executive Samuel Palmisano said that this year it'll sell a workstation containing Cell. That's true, it will be built at IBM's Boeblingen facility near Stuttgart (see Tech Watch, November 29). Such a system will be used for software development for the future Sony console. It's possible it may be used for scientific work, but special compilers and applications will have to be created to make use of it. That comes under the "someday" heading.
Is it a superomputer, on a chip?
Well yes, architecturally, it's exactly like the multi-processor SIMD vector machines that were built in the mid-eighties, and to some extent are still being built (only 4.2% of today's supercomputers are vector based). However, at 256 GFLOPS it's not, by itself, in the supercomputer class. (The slowest machine on the current list of the Top 500 supercomputers does 624 GFLOPS http://www.top500.org/lists/plists.p...5&M=06&Y=2004.) But you can envision an array of Cells being made into a supercomputer, and given the expected low cost of this soon-to-be-mass-produced chip, that's entirely feasible, except that it might wrinkle the feathers of IBM's BlueGene group a bit.
It's not a Grid computer
The other popular misconception, which IBM and Sony are doing nothing to dispel, is the notion that the Cell will hook up to some type of grid (like Butterfly net) and either tap into more resources or allow millions of user to simultaneously play a game. Yes, online gaming can take advantage of the Butterfly net, and yes, IBM and Sony said they would use it. But it's not to share Cell resources?it's to play against others just like any other online network.
Take a look at this quote from a story published by CNET:
"Cells can even roam over a network, allowing the processor to perform a type of distributed or grid computing, an increasingly popular enterprise technique in which demanding tasks are divvied up among a gang of networked computers."
Clearly, there are some major misunderstandings about what networked and Grid computing are all about. Grid is used for collaboration and uses the Internet, not the fastest data exchange network in the world and certainly not fast enough for distributed for real-time computer, the type of computing that epitomizes a game.
A Playstation 3 could borrow unused processing power from other consoles on a network, for example, to complete a demanding task such as delivering streaming video. Sony has done a masterful job at creating a science fiction scenario that ignorant journalists have latched onto. The only problem is some of those ignorant journalists have good memories, and in a year or two when the next-generation Playstation is delivered and it doesn't do all these suggested whiz-bang things, those same journalists are going to feel used, and then they will write searing stories about how Sony and IBM have failed to live up to the expectations and promise of the lauded Cell. This misdirected PR BS is going to haunt IBM and Sony. Lest we forget, when Sony introduced the Playstation 2 Emotion Engine it was promised to power a wide variety of consumer electronics, and so far it's failed to live up to that promise.
Is it your next TV?
Well, Toshiba has been quoted as saying it plans to incorporate it into high-end TVs. Given the crashing ASPs on TVs these days, it's not likely the Cell, in its current configuration, will show up inside a TV, high-end or otherwise. The vector processors could be useful for MPEG decoding/encoding, but you wouldn't need eight of them, so a stripped-down version, with maybe two SPEs and a Power CPU, might be made for high-end TVs sometime in 2006, maybe.
So what is a Cell?
It is an eight-IMD processor array with a powerful master controller and program execution GP CPU. Each SIMD has local memory, high-speed interconnects, and a segmented memory structure ideally suited for graphics and codec applications. It has a super-high-speed chip memory interface and an equally fast flexible I/O for the graphics processor. It is currently running Linux and could run other OS's on the GP CPU.
It will be a really great game console processor and general multimedia accelerator, and should cost out at a reasonable consumer price range.?Jon Peddie
Originally posted by cben
link below helps dispell some of the misconceptions about Cell. It doesn't sound like Cell is really as applicable to the PC as we thought or at least according to this guy.
http://www.jonpeddie.com/index.shtml
P.S. I don't know that much about computers so be please be easy on me.
Ha! This guy (follow the link) says:
The Cell is not a Pentium killer or a competitor?it's a SIMD supercomputer, and just because it has an IBM Power processor in it, which has been demonstrated to run X86 code in an emulation mode, that's not the use or the future of this Cell.
...
The Cell has an array of SIMD RISC processors, not multi-cell X86 processors.
He understands Intel competion = Pentium emiulation
Originally posted by shadow
Isn't Sony supposed to use the chip for gaaming?
As BenRoethig pointed out, I wasn't referring to Cell's theoretical capability. Obviously, anything going in a PlayStation has to be suitable for games.
But nearly all the A-list games for the Mac are ported over from Windows, which means they're designed and implemented to run well on Windows, which means they're single-threaded applications optimized to run on one powerful, scalar core, which means they'll run poorly on Cell (especially relative to its theoretical capability). Any optimization for Cell would be up to the porting houses, and it might be impractical or even impossible given the codebase, and time and money constraints, and other issues.
Therefor, practically speaking, Cell would not bring much to the Mac as far as mainstream PC gaming is concerned.
I wouldn't hope for ports from the PS3, either. A Mac game does well if it sells in the tens of thousands. For most console games, that's margin-of-error range, not worth the trouble and cost. In fact, if the threads I've read are any indication, you'd actually target more Mac gamers with the console version...
Apple invested a lot in AltiVec optimization. The Core Data\\Image\\Video... collection suits data parallelism very well. And (it seems) they are moving all their multimedia software and iApps to those frameworks. That makes me think that they have the Cell in mind.
Possibly, but not necessarily. After all, all that stuff runs really well on the GPU. Apple can afford to wait and see, or have IBM make them a custom configuration.
Originally posted by mikenap
wow, powerful article about cell and Apples potential in using it. an intereting read, please post feedback...
http://www.blachford.info/computer/Cells/Cell4.html
Okay, I'm way behind in this thread (as well as my work) and will try and catch up as time permits, but one thing that hits out to me in this article is this quote,
The Core Image technology due to appear in OS X "Tiger" already uses GPUs (Graphics Processor Units) for things other than 3D computations and this same technology could be retargeted at the Cell's APUs. Perhaps that's why it was there in the first place...
The barrier to CELL discussed earlier in this thread and which is the same old problem going forward with new tech, is the need to re-write apps, modify the OS, etc, all of which can be a painful, arduous and expensive process.
Could Core Audio, Core Image, and Core Data already be ready to be offloaded to different cells. I'm just a pixel pusher, but it seems to me that this would be feasible with current apps working fine with the core PPC 970 + Altivec and then the developer implementing over time Core Audio, Core Image, and Core Data API's without much of a hiccup?
Could this be part of Apple's plan all along or am I way of base?
Sorry, if someone down stream has already commented about this.
Originally posted by tink
Okay, I'm way behind in this thread (as well as my work) and will try and catch up as time permits, but one thing that hits out to me in this article is this quote,
The barrier to CELL discussed earlier in this thread and which is the same old problem going forward with new tech, is the need to re-write apps, modify the OS, etc, all of which can be a painful, arduous and expensive process.
Could Core Audio, Core Image, and Core Data already be ready to be offloaded to different cells. I'm just a pixel pusher, but it seems to me that this would be feasible with current apps working fine with the core PPC 970 + Altivec and then the developer implementing over time Core Audio, Core Image, and Core Data API's without much of a hiccup?
Could this be part of Apple's plan all along or am I way of base?
Sorry, if someone down stream has already commented about this.
To a pixel pusher. Apples plan was, and is to offload 2D, and 3D graphics work to the GPU using core image API. It's all been stated at WWDC 2004 (video still available at Apple.com/quicktime), and in developer documentation. Using any new tech like this usually extremely simple being that it's probably part of a system -.framework which for the most part is like drag, and drop mach O, and OO programming.
That was my interpretation of Apples presentation from WWDC, but feel free to correct any mistakes.
Whatever happened to those trinity Rumors....
cell = Trinity (3 companies IBM Sony Toshiba)
http://forums.appleinsider.com/showt...rmac+and+pixar
and what about this thread....
http://forums.appleinsider.com/showt...es+of+the+past
given cell is going to be in the playstation then Dolphin ....
This thread actually could have or still is prophetic...!1
Amorph, you were there, help me out!
8)
Originally posted by onlooker
To a pixel pusher. Apples plan was, and is to offload 2D, and 3D graphics work to the GPU using core image API. It's all been stated at WWDC 2004 (video still available at Apple.com/quicktime), and in developer documentation. Using any new tech like this usually extremely simple being that it's probably part of a system -.framework which for the most part is like drag, and drop mach O, and OO programming.
That was my interpretation of Apples presentation from WWDC, but feel free to correct any mistakes.
I am well aware of Apples plans and I understand the intent to offload 2D and 3D graphics to the GPU using the Core Image API, however my point was based off the comment by the author of the article where he speculated that
this same technology could be retargeted at the Cell's APUs.
The point being that to utilize CELL technology would be infact relatively painless for the same reason you point out.
Using any new tech like this usually extremely simple being that it's probably part of a system -.framework which for the most part is like drag, and drop mach O, and OO programming.
One reason not to use CELL often assumed would be the huge amount of effort it would take to rewrite everything. The authors point that the Core frameworks could be "retargeted" to use CELL instead of the GPU is what I found interesting.
As I mentioned before, I'm just a pixel pusher and what I thought was an interesting read was pointed out by Programer to be by an author that was full of sh. If that's actually the case or if Programer woke up on the wrong side of the desk, I don't know.
Originally posted by tink
One reason not to use CELL often assumed would be the huge amount of effort it would take to rewrite everything. The authors point that the Core frameworks could be "retargeted" to use CELL instead of the GPU is what I found interesting.
Well yes the author is a total dreamer if he thinks it's going to happen suddenly, or maybe he is speaking of a future cell processor. But, that would be the I/O's for the GPU to presumably offload operations he is referring to. I don't think it's necessary. A good nVidia GPU wont need it, but lesser GPU could conceivably use it, but not this one any time soon.
This particular cell is not going into a Mac ever. When, and if one eventually did I think by the time it does it will be completely different, and evolved from what it is now. - Which is basically a diagram/structure to base the PS3 on. The author is stirring a muck of assumption, presumption, and not interjecting the most critical points to his hypothesis, or he is just taking bit's, and pieces of what everyone else says, and spitting them all out at once like they all belong together, and make sense now without any of the foundation material. But if I say much more about it I'll be doing the same thing because I have not seen his article, only what you copied.
Originally posted by TednDi
OK kids file this one under "The memory HOLE"
Whatever happened to those trinity Rumors....
and what about this thread....
http://forums.appleinsider.com/showt...es+of+the+past
given cell is going to be in the playstation then Dolphin ....
This thread actually could have or still is prophetic...!1
Amorph, you were there, help me out!
8)
oh my god i am tripping. just like that nostrus dude!