Catch up on stories from the past week (and beyond) at the Slashdot story archive

 


Forgot your password?
Close
typodupeerror
×
IBM PlayStation (Games) Hardware Games

IBM's Plans For the Cell Processor 124

Posted by Soulskill from the breeding-a-better-hamster dept.
angry tapir writes "Development around the original Cell processor hasn't stalled, and IBM will continue to develop chips and supply hardware for future gaming consoles, a company executive said. IBM is working with gaming machine vendors including Nintendo and Sony, said Jai Menon, CTO of IBM's Systems and Technology Group, during an interview Thursday. 'We want to stay in the business, we intend to stay in the business,' he said. IBM confirmed in a statement that it continues to manufacture the Cell processor for use by Sony in its PlayStation 3. IBM also will continue to invest in Cell as part of its hybrid and multicore chip strategy, Menon said."
This discussion has been archived. No new comments can be posted.

IBM's Plans For the Cell Processor More

IBM's Plans For the Cell Processor

Comments Filter:

  • We like money! (Score:4, Insightful)

    by allaunjsilverfox2 ( 882195 ) on Tuesday October 12, 2010 @01:41AM (#33866436) Homepage Journal
    What business would want to give up guaranteed sales? I mean, a gaming platform is like walking into a bank, depositing one cent and then getting a cent every second until the bank closes.

  • Multicore for raytracing? (Score:3, Insightful)

    by gizmod ( 931775 ) on Tuesday October 12, 2010 @01:41AM (#33866438)
    Bring on a 12 core PS4 with raytracing games.

    • Re: (Score:2, Informative)

      by Anonymous Coward

      The Atari Transputer Workstation [wikipedia.org] already did that in the 80s. Coolest real-time raytracing ever!

      • Hmm. I'd never looked into the Atari Transputer much. I figured it was a lot like an Amiga 2000/3000, but overhyped, and with GEM :) Turns out it was quite a machine, with a lot of innovation that's only catching on in PCs now. If it wasn't for the lack of an MMU, I might have liked to see it replace both Amigas and PCs :) Also, a lot of the stuff here:

        http://en.wikipedia.org/wiki/Transputer [wikipedia.org]

        Sounds like a summary of the Cell's raison d'etre.

        Couple of questions:

        * Is there an emulator of this, so I can ch

        • Re: (Score:1)

          by AlecC ( 512609 )

          There is an emulator, under active development, See posts on newsgroup com.sys.transputer.

          The 260 megaflops must be for some kind of an array - they were designed to be used in arrays. The individual transputers never clocked faster than 25 MHz, though the FPU on the T800 was relatively fast for the time. Each transputer had four bidirectional links connected to DMA engines wired directly into the hardware scheduler, so that inter-processor communications were very low cost.

          I can't see why the architecture

      • And of course Pixars little foray into world of computer hardware, the RM1. Not quite a raytraced renderer as such (REYES), but it was actually used for commercial films very briefly (e.g. TinToy, StarTrek2).

        These days you fire off renders by invoking the prman executable.... 'p' being short for 'prototype' (it became quickly apparent to Pixar that Sgi's development and performance curve was outpacing their own hardware division. Rather than try to compete, they simply shut down the hardware division,
    • No please don't. We've got enough on our plate with the PS3 as it is.....
  • I wish I could buy a consumer-priced system with one of these CPUs. A very interesting system to develop for. After all, we all are going to use some kind of system with the separate memory model, more like this, when we will come to the end of scalability of the currently dominating multicore CPU with common memory space.

    I hope that PS4 (or other console using it) will be linux-friendly as PS3 was until Sony blew it. Alas, however slim this chance is, there seem to be no better chance.

    • Re: (Score:2, Informative)

      by Nursie ( 632944 )

      Get PS3 with 3.41 or earlier firmware -> Jailbreak -> install linux.

      Profit?

      (actually linux for jailbroken ps3s is in the very early stages, but I'm sure it'll get there.

      • ...platform where you don't have to worry about some idiot company dictating what software you run on the hardware you purchased, don't you think?

        • Re: (Score:2)

          by Nursie ( 632944 )

          Oh sure, but if your aim is to try out cell programming, then that's pretty much your only option at present!

          It would probably be better to try using CUDA and your graphics card...

  • game over more cores with less heat.
    Great, but where is the software expert side going to come from?
    It seems to take years for any 3rd party to work out how to optimise "anything" HD for the systems.
    With a push for more cores how about a push for more developer support vs "cloud-based" and p2p servers.

  • With more memory per CPU, it might not suck (Score:5, Interesting)

    by Animats ( 122034 ) on Tuesday October 12, 2010 @01:54AM (#33866498) Homepage

    The basic problem with the Cell processor is that it has 256KB (not MB, KB) per processor, plus a bulk transfer mechanism to main memory. Given that model, it has to be programmed like a DSP - very little state, processing works on data streams. For games, this sucks. No CPU has enough memory for a full frame, or for the geometry, or a level map. Trying to hammer programs into that model is painful. (Except for audio. It's great for audio.) In many PS3 games, the main MIPS machine is doing most of the work, with the Cell CPUs handling audio, networking, and I/O. And, of course, Sony had to put an NVidia graphics processor in the thing late in the development cycle, once people finally realized that the Cell CPUs couldn't handle the rendering.

    But if each Cell CPU had, say, 16MB, the Cell machines could be treated more like a cluster. Programming for clusters is well understood, and not too tough.

    It's probably too late, though. Multi-core shared memory cache-consistent machines are now too good. It's not necessary to use an architecture as painful as the Cell. It's probably destined for the graveyard of weird architectures, along with data flow machines, hypercubes, SIMD machines, systolic processors, semi-shared-memory multiprocessors, and similar hardware that's straightforward to build but tough to program.

    • Well, IIRC it's intended purpose was for embedded devices. They were talking about smart fridges, security systems, etc. Basically networking your home with smart devices that were running on Cell and then being able to use that processor juice distributed across the devices since the Cell scales very well(which is why the PS3 makes a great supercomputer farm)

      • Re: (Score:1, Informative)

        by Anonymous Coward
        No; no you don't recall correctly, not even a little bit. Not a jot, not a tittle. Cell was designed specifically for the PS3, and maybe for other kinds of (repetitive streaming type) work that is mostly done by GPUs and/or CUDA in this day and age.
    • Yep, Cell is being used far outside its original design spec. Of course, if gaming consoles is its current largest market, the next generation will probably look much more like a standard POWER6 or 7 in its architecture - more emphasis on more powerful support cores, more memory per core, and all the other things that have made their way into every other CPU family currently popular.

      • Back in the early PS2 we would talk about what a next generation PS2 would look like. Those whiteboard diagrams looked almost identical to what Sony and IBM came up with.

        The parallels between the PS2/EE/GS and PS3/Cell/RSX are almost identical:

        Execution starts on the EE/PPU
        Heavy/parallel computation task is spawned off to the VUs/SPUs
        Light control code runs in parallel on the EE/PPU
        As graphical elements become read to be rasterized they are spawned off to the GS/RSX

        In a well running PS2/PS3 engine all three

        • Re: (Score:3, Interesting)

          by CronoCloud ( 590650 )

          I remember reading somewhere that one of the goals in PS2 programming was keeping that DMAC running full tilt streaming data. Ah, found it, Ars Technica:

          http://arstechnica.com/old/content/2000/04/ps2vspc.ars/4 [arstechnica.com]

        • I found this article interesting. They write about Valves approach to multi-core CPU's and game engines.

          The programmers at Valve considered three different models to solve their problem. The first was called "coarse threading" and was the easiest to implement. Many companies are already using coarse threading to improve their games for multiple core systems. The idea is to put whole subsystems on separate cores; for example, graphics rendering on one, AI on another, sound on a third, and so on. The problem

    • ...processor is that the company selling it's flagship product decided to lock out people wanting to experiment with it.

      • Re: (Score:1, Insightful)

        by Anonymous Coward
        ...processor is that the company selling it's flagship product decided to lock out people wanting to experiment with it.

        Because those people made such progress after having nearly four years to experiment? It's time people around here quit pretending like Sony never gave them the chance to dink around with the PS3.

      • ...processor is that the company selling it's flagship product decided to lock out people wanting to experiment with it.

        Fail. The flagship Cell processor is a more-capable unit that IBM will sell you for exorbitant amounts of money. The Cell in the PS3 is a toy version and even mentioning that it is based on cell is only marketing for the real thing to IBM.

        • I bet you didn't like that I used the word "Fail" in the modern vernacular sense. But in case you thought I was being non-factual, here is information on the real cell processor [wikipedia.org] which IBM sells for truly incredible amounts of money. I've looked up the pricing, and it is scary.

      • Re: (Score:1, Troll)

        by Lunix Nutcase ( 1092239 )

        And nothing of value was lost to them. The only thing related to the PS3 that interests Sony is the selling of games, Blu-Rays and stuff from PSN. A bunch of basement dwellers installing Linux on their PS3 was an afterthought at best.

    • Laughable Drivel (Score:2, Interesting)

      by RingBus ( 1912660 )

      "Sony had to put an NVidia graphics processor in the thing late in the development cycle, once people finally realized that the Cell CPUs couldn't handle the rendering."

      My god. You are repeating that Beyond3d forum lie in late 2010???

      "For games, this sucks"
      "Trying to hammer programs into that model is painful. (Except for audio. It's great for audio."
      "In many PS3 games, the main MIPS machine is doing most of the work, with the Cell CPUs handling audio, networking, and I/O."
      "It's not necessary to use an arch

    • The bandwidth in and out of those tiny spu memories is great, much better than between main memory and cache on an x86 processor, or generally between cache and processor on a GPU. I don't know what anyone needs that for though.

    • Re: (Score:3, Interesting)

      by CronoCloud ( 590650 )

      The SPE's aren't full CPU's, they're essentially enhanced versions of the PS2's VU's.

      Given that model, it has to be programmed like a DSP - very little state, processing works on data streams.

      Yep, stream data, just like on the PS2.

      For games, this sucks. No CPU has enough memory for a full frame, or for the geometry, or a level map.

      You're not supposed to keep a full frame or map in there, you're supposed to stream it in and out on the fly, as the Kami intended, just like on the PS2.
      "Fat Pipes (bandwidth), sm

    • Please don't confuse the SPUs (the eight coprocessors on the Cell die) with the PPU (the main CPU core). The PPU is also part of the Cell, so don't call the SPUs "Cell CPUs". There is also no MIPS core -- the PPU is a 3.2GHz PPC core with two hardware threads. The SPUs also run at 3.2GHz, but are not considered "real" CPUs since they can't bootstrap themselves, they have to be given tasks from the PPU. SPU programming forces a model on you as a developer -- modularize your tasks with as few synchronizat
  • I would not want to be betting against IBM for this marketspace. Their cell chip, which is an asymmetric multi-core CPU architecture, seemed bizarre when announced, but has proven to be quite good for these workloads. If IBM is looking to leverage their regular POWER chipset for the console market, they will probably build some screamers with them. Cell and POWER both have Unix and Linux adaptations running on them, so having the capability seems trivial. Whether vendors will want you using their hardwa

    • Re: (Score:1, Troll)

      by TheRaven64 ( 641858 )

      If IBM is looking to leverage their regular POWER chipset for the console market, they will probably build some screamers with them

      All of the current generation consoles use IBM chips. The GameCube and Wii both used PowerPC 4xx series chips - IBM's low-end 32-bit PowerPC line. The XBox 360 uses a custom 3-core in-order PowerPC chip. The PS3 uses Cell (PowerPC core + 7 SPUs - the PS3 gets the ones where one of the SPUs failed the tests, the ones where all 8 work go into blades and supercomputers).

  • They could actually try to sell the Cell (Score:5, Interesting)

    by dbIII ( 701233 ) on Tuesday October 12, 2010 @02:00AM (#33866522)
    A while back I was looking for one or two Cell CPU based machines as development boxes for inhouse geophysical software - basicly to see if it's worth going onto that platform. The three week process between contacting what appeared to be the only vendor of Cell based workstations and getting a price for an entry level machine was frustrating. It involved daily calls to a slimy bastard that appeared to just want to waste time trying to become my friend until he had carefully finished weighing my companies wallet.
    In the end the time window had come and gone (the developers got bored or gave up on the idea of using the Cell) before I could get even a hint at the price but I kept going for the sake of future projects. The price for one workstation with one processor was fairly similar to that of six of our cluster nodes. You would need some sort of black-ops budget where any Accountants coming close are shot on sight before paying that sort of price. An entry point machine no much different to a playstation with more memory cost a truly insane and unjustifiable price.

    • That is true. I don't know about now but a few years ago you couldn't even get a pinout for the Cell processor. You had to show both IBM and Sony your business plan and your market could not impact Sony. IBM has some cookie-cutter circuit boards with a cell on them that they want to sell for big bucks, along with a big down payment and minimum quantities. The reality is that the Cell processor is not THAT great. good, but not great, and requiring a big change in the way you factor out your software design,

    • Re: (Score:3, Informative)

      by TonyMillion ( 545370 )

      odd, when we were working with cell we went straight to matrix vision and they LOANED us the hardware for about a year.. Nothing sleazy at all. IBM Also loaned us a server, as did Sony (a beautiful rack-mount job which will never see the light of day).

      http://www.matrix-vision.com/products/cell.php?lang=en [matrix-vision.com]

      Bottom Line - the PPC part of the Cell is rubbish, terrible IO and generally 'weak' by todays standards, the SPEs are great, but not enough memory on them (256k) for the algorithms + tables we needed to pro

    • Re:They could actually try to sell the Cell (Score:5, Interesting)

      by wowbagger ( 69688 ) on Tuesday October 12, 2010 @07:30AM (#33867796) Homepage Journal

      I can go one better: I do signal processing for a living - chewing on multi-hundred megasample/second streams of data in real time. The Cell looked like a perfect fit. We were looking at 1000's per year. Contacted IBM - sorry, not enough zeros on that number for us to sell you the chips. OK, are there any vendors that are targeting the uTCA form factor (that the Telecomms folks are are all over, so they would not have been targeting just us)? Nope, just large blades for mainframes.

      I assert that IBM doesn't want to be in the chip business - at least, not "selling chips to anybody else". They don't mind making chips for their own use, but they really don't have the infrastructure to sell to anybody else.

      Sony and Toshiba don't want to be in the high-end CPU market, they want to be in the mass-market stuff.

      Had IBM licensed the Cell design to somebody like Freescale, they might have gone somewhere.

      Sorry, but I RTFA - and what I came away with was "We will continue to support Sony for as long as Sony wants to make PS3's". I saw nothing that really said "We are going to be going someplace else with this."

  • The problem with Cell... (Score:3, Insightful)

    by Entropius ( 188861 ) on Tuesday October 12, 2010 @03:50AM (#33866952)

    ... is that it lies in between ordinary x86-type multicore processors and CUDA/GPGPU, and there's not much room in between.

    • Re: (Score:1, Troll)

      by TheRaven64 ( 641858 )

      It's more of a triangle. In one corner, you have general-purpose CPUs, optimised for branch-heavy code with lots of locality of reference. In another, you have streaming, often SIMD, processors optimised for non-branching code, with high throughput, such as GPUs and DSPs. In the third corner, you have specialised silicon dedicated to specific algorithms (e.g. building blocks for encryption algorithms or video CODECs).

      Cell is along one side of this. It isn't particularly throughput-focussed, and it is

  • I know Slashdot is the enemy of good writing practices, so this post will be modded downto hell, but I feel I must point out that lately, the capitalization of titles of Slashdot submissions got completely out of hand. The rule is simple: if you want to capitalize your headlines, you capitalize every word except
    - prepositions ("of", "to", "in", "for", "with" and "on")
    - articles ("the, "a" and "an")
    - and some other obvious exceptions.

    On Slashdot, the editors are so ignorant that they usually capitalize each

  • I wonder what it's gonna have to absorb to evolve into the Perfect Cell. :)

  • what would be cool (Score:3, Interesting)

    by AVryhof ( 142320 ) <amos@NospaM.vryhofresearch.com> on Tuesday October 12, 2010 @06:50AM (#33867610) Homepage

    What would be a pretty cool chip would be an 8-core chip with 4 x86_64 cores, two graphics cores, and two Cell cores. (perhaps IBM + AMD working together)

    After that, build a custom Linux with MeeGo as the front end / launcher.... It would be cool if game console makers embraced Open Source for everything up to launching the games. ...and if they don't want their SDK open source, that's fine, just make the Operating System so it can launch the games, then get out of the way. Run it on two cores (for better functionality with Multimedia capabilities, ebook reading, etc.) and use the rest of the cores (2 x86_64, 2 Graphics and 2 Cells) for gaming.

    As for the other hardware, Composite, Component, HDML, VGA, WiFi, Ethernet, and a headphone jack.(maybe bluetooth for wireless controllers and the ability to use bluetooth headsets)..blu-ray, card reader, and USB.

    This is all off the top of my head, and would be a pretty cool gaming console, which would truly capture the home entertainment medium and make most people looking for gadgets, consoles, or HTPCs drool appropriately.

    • What would be a pretty cool chip would be an 8-core chip with 4 x86_64 cores, two graphics cores, and two Cell cores. (perhaps IBM + AMD working together)

      this is a bad idea because it's precisely the kind of ignorant crap that hypertransport is supposed to eliminate. Instead of cramming a bunch of crap into one package, you sell multiple packages so that people can customize their layout. Ideally you'd have the x86_64, cells, and graphics cores all communicating via HT links, and then it doesn't matter where they are physically located... but trying to put all that into one package would be a TDP nightmare at this point.

  • I've wondered this for a while: IBM, Toshiba, and Sony developed the cell. Pray tell, they sell a cell, how do they divide the income?

Slashdot Top Deals

Our OS who art in CPU, UNIX be thy name. Thy programs run, thy syscalls done, In kernel as it is in user!

Close