IBM Plans to Open the Cell Processor

morcheeba writes "According to an EETimes article, IBM is planning on releasing the full specifications and software libraries for the powerful processor that will be in the Playstation 3. The goal is to stimulate open-source development for other applications of the chip. The article doesn't mention if there will be some affordable development systems for all these programmers -- I'm hoping for a ps3 devkit." From the article: "IBM is eager to find other opportunities for Cell, but it's going to take a lot of software work...Going to the open-source community makes sense, because they could attract a lot of pretty smart programmers who could spin out software and applications for Cell."

All 3 consoles = IBM?

[by Anonymous Coward]

I just realized this means both next-gen consoles and the GameCube all use IBM processors. Impressive. Now I hope they can hang on to the Mac market, and maybe both will benefit from advances in the other.

What else?

[mnmn (145599)]

What else is out there anyway? The main contenders are PPC and Intel. Both Intel and AMD produce x86/64 chips only. Sure everyone also produces arm/strongarm chips, but theyre still weak, the fastest strongarm from Intel is used on higher end PDAs.

Whats left is MIPS, Ultrasparc, PA-RISC, Alpha and special purpose FPGA chips.

MIPS is dead. SGI was producing servers on Itanium which also died.

Ultrasparc is dying in favor of x64. Sun guards its IP jealously. Low throughput but high floating and thread performance.

PA-RISC gives the best bang for the MHz. Good float, everything else runs too hot for now. Old old architecture.

Alpha was killed by HP. They'll try to sell you Itanium or PARISC before they sell you an Alpha. Development on it has completely stopped since 21264c. And I mean COMPLETELY.

FPGA chips are less efficient, and better use an ARM than an fpga chip.

So the two champions are PPC (and its derivative, Cell) and x86/x64.

Architecturally, PPC, and a 64-bit-only x64 are efficient. But IBM has been trying to push PPC in the market, working hard on a grand plan to take the market dominance away from x86. Look at all their offerings for Linux on PPC. They're prepping up this combination against wintel... and any usage of PPC means profits for them and Motorola, mostly to IBM in the higher end.

The choice is rather easy. If you will not use an IBM chip for a higher-end game console, what will you choose?

Sam Fisher?

[Mille Mots (865955)]

I guess if they encourage development outside of the Sony realm, they'd be fostering a 'Splinter Cell?'

Calling all pawns...

[PenchantToLurk (694161)]

IBM wants open source dev on cells like MS wants developers coding for Windows. It's an sales tool to convince manufacturers to source their parts.

Given that the only cell device is the PS3, and that sony would sooner slit their wrists than let users write their own code for it, we can only assume that IBM is hoping somebody else will pick up the cell for consumer devices.

The next "IBM Compatible" standard?

[Rude Turnip (49495)]

The original IBM compatible standard (what we just call X86 nowadays) took off when Compaq reverse-engineered the BIOS and created the first "clone" of a "genuine" IBM PC. This undoubtedly resulted in explosive growth for Intel, who made the CPU. Now that IBM is manufacturing the chip (instead of Intel for the first IBM PC), it is absolutely in their best interest to make the Cell processor as mainstream as possible by opening up all of its specs.

Linux anyone?

[sleepingsquirrel (587025)]

Linux Insider is running a couple of editorials speculating [linuxinsider.com] about running Linux [linuxinsider.com] on the 'Cell'. The bold prediction? 'the Linux developer community will, virtually en masse, abandon the x86 in favor of the new machine.'

Re:Linux anyone?

[zdzichu (100333)]

Already done! Linux supports Cell processor [lkml.org] since May 13.

Functional Compilers, anyone?

[jameson (54982)]

Good thinking, IBM. Now, let's get SML/NJ [smlnj.org], Haskell [haskell.org], and O'Caml [ocaml.org] ported to these things.

"Why", you may wonder, but the answer is simple: Referential transparency or any kind of confinement of side-effects makes for easy parallelisation, which is what these Cell thingies are supposed to rock at.

This might be the one thing that will put FP back into the undergraduate curriculum.

-- Christoph

Re:Functional Compilers, anyone?

[Wesley Felter (138342)]

The reason why I mentioned "putting FP back into the curriculum" was, however, that it is my understanding that, if you're right, there's a good chance that programmers would prefer multithreading in imperative languages precisely because it'd be closer to what they'd be used to. So, by getting them used to FP, we'd see a "more fair" evaluation of the practicality of this approach.

Your idea seems to be that idealism would drive the CS curriculum, which then would drive the industry. My observation over the last 10 years has been the opposite. In 1996, CS freshman were learning Scheme and Haskell; today they are using Java, because "it's more practical and aligned with the industry" or some such excuse. But now that the bust has eliminated all the "I just want to get rich" CS students, maybe it will swing the other way.

Open HARDWARE movement

[WebCowboy (196209)]

Nice to see someone as important as IBM realising the importnace of open HARDWARE. I've found that until recently the concept has been overlooked or even derided. Even open software advocates didn't "get it" and said it could never work, becasue hardware is different--the argumant was that hardware isn't something individuals or small companies could influence becasue of the high cost of entry, and big companies needed to make money off licensing closed IP to fund development and production of new hardware.

This day and age, such an argument is complete BUNK. Hardware design is done on computers and chip specifications are more often than not specified in VHDL or Verilog--the "source code" of hardware if you will. Not only is design and simulation within the reach of even hobbyists, the end result is very similar to software in characteristics. While IBM is not completely opening things up to the point of showing the "source code" of the Cell processor, it is a great step to see all the specifications etc. without encumbrances.

Quite frankly I'm surprised the open source movement hasn't advocated open hardware much more vigourously. After experiences around NVidia and ATI and Intel you'd have to be a fool not to realise that open hardware isn't just an interesing idea, it is NEEDED for the success of open software.

OpenCores

[jd (1658)]

Open Cores [opencores.org] is a site that aims to do for hardware what Sourceforge did for software. They have a good collection of processors, memory chips, network and other I/O devices, etc.

I've got my own project on there, in a bid to develop a totally parallel OO-based processor, but not had much time to work on that recently.

Those interested in Open Hardware should visit this and similar sites, to see what is happening out there, whether or not they believe the idea could work in practice. Why? Because it is an excellent source of ideas, and ideas are what keep all the IT markets moving.

A free anti XBOX 360 Army

[Irish_Samurai (224931)]

Why wouldn't Sony want IBM to do this. Since the platform specific game developer is going the way of the Dodo, how do you get an edge?

Perhaps by giving every anti-Microsoft fanatic video game freak an outlet? When licensed 3rd party support becomes even on both sides of the map, it will be consumer mods that make the difference to gamers. Can I mod chip it to play foreign games? Can I put vinyl kits on it? Can I use it to power my toaster?

Theoretically, one might be able to write some code that will allow you to play foreign games without having to void your warranty. How huge is that?

Also, Sony is going to need something extra to get people to buy it's system after a XBOX 360 Holiday season, and this may just be that.

Fortune Cookie in the Oven

[Nom du Keyboard (633989)]

"It can be hard to give stuff away for free," Kahle said in an interview

This has the sound of the next Slashdot Fortune Cookie in the making -- or should I say in the baking?

Script Flip Chip

[Doc Ruby (173196)]

This setup looks a lot like the inverse of IBM's greatest strategic loss: the PC. That time, IBM had the brand, and sold hardware to a maverick niche market of PC hobbyists. They viewed software, including the OS, as a necessary sidelight, and let Microsoft judo them out of their control of the market revolution they created. IBM later lost $20B as their market failure came home to roost, and never recovered the leadership they squandered.

This time, IBM is the necessary part in the Playstation, which is in the hands of this generation's maverick niche market: gamers. Their Cell processors give them Microsoft's opportunity: base the market in the demanding niche, and market their product outside of it, leveraging their market feedback and brand into the larger market, including supplying competitors to the original platform. IBM is flipping the script: selling hardware means opening the software promotes their sales, inverting Microsoft's formula of taking software proprietary to capture more of the market defined by the hardware.

It all looks great on paper. Especially the greater scalability and persistence of open software, compared to Microsoft's centralized, proprietary approach. Time will tell if IBM can manage the opportunity, competing against Microsoft, as well as Microsoft did in the 1980s - and better than Microsoft will in the 2000s.

Steps for Opening Cell

[mnmn (145599)]

(1) Port GCC to it, optionally another much more optimized compiler that is compatible with gcc.
(2) Give it to taiwanese motherboard makers to make microatx mobos on the cheap. Aim for $40 for lower speed ones and $100 for full speed Cells.
(3) Put out all the specs of the Cell and any possible firmware sources online, and put them under the BSD license.
(4) Provide licenses to other devleopers to make cheaper versions of the Cell.
(5) Watch Linux and NetBSD grow on it. Watch cisco use it on their high-throughput routers and other manufacturers use it. Watch the app base grow.
(6) Profit!

Alternatively sit on it and let it rot like Palm is doing with BeOS.

Re:What I wonder...

[LWATCDR (28044)]

Probably do not care. The Cell is only part of the PS3 it is pretty unlikely that anyone else will build a console with it. It could actually help Sony in the long run.
The more people that buy Cells to put in to widgets the lower the cost for Sony.

Re:What I wonder...

[Wandering-Seraph (878056)]

"In this light, IBM, Toshiba and the third Cell partner, Sony Corp., are turning to the open-source community to drum up interest in the architecture."

Doesn't sound like they're particularly upset about it.

Sony must have approved this

[John Seminal (698722)]

how does Sony feel about this?

Sony must have given its approval for releasing this information. It could not happen without them.

If Sony did not know, and IBM made this move without their approval, I could see Sony NEVER buying from IBM again. That is too big a risk for IBM. Heck, most companies would think twice.

Will it be easier to make a mod chip if people know how the processor works? Or did Sony add their DRM elsewhere? Who knows. IBM is not releasing the blueprints for the Playstation 3, just the processor.

Sony is a big company that hires smart people. Maybe they figured out hiding the electronics will not prevent reverse engineering. Maybe the new PS3 has some technology that makes it difficult to mod.

Maybe this is like Microsofts WMV, it is unhackable, nobody can get it to play a stream if DRM v9 is enabled. Not one person on the planet. And it has been over a year now.

For the PS3, they don't need for their game machine to be unhackable forever, just until the PS4 comes out. :)

Re:Sony must have approved this

[Alsee (515537)]

The DRM enforcement is inside the CPU itself. Trying to add a mod-chip likely isn't going to work short of replacing the CPU itself. Replacing it with another CELL chip just stick you back in the original DRM jail cell, and trying to replace it with another CPU or with an emluated CELL won't work.

Each Cell is given a GUID, a global identifier, [theregister.co.uk] and will come with a crypto-signature authenticating it as a genuine DRM secure chip.

You can't defeat the system without (1) extracting secret keys from each chip one-by-one, or (2) generating a fake crypto signature to falsely authenticate a non-DRM enforcing chip. If you do manage to extract a key from one of the chips and they find out about it, they will place that key on a revokation list and it will become useless. So each chip you manage to rip and extract a key is good for creating one "liberated" system, and you still have to be extremely careful that no one can ever detect that you have done so.

The Pentium 3 unique ID numbers got killed off because of public outrage, and that system was nothing compared to what they've built into the CELL processor. It's about time we see some coverage of this aspect of the chip, and refuse to buy any CELL chips or CELL hardware unless these UNIQUE PROCESSOR IDENTIFIERS are removed.

-

Re:Neat But...

[Nom du Keyboard (633989)]

How would someone go about acquiring a CELL chip and other hardware required to run it?

Buy a PS3?

Cell: new desktop processor, or video-card killer?

[wiml (883109)]

Linux? Sure. The "PPE" portion of the Cell is a POWER64, which Linux already runs on. The "SPE" engines are effectively going to need their own kind of OS to manage them, but you could start with a mostly-user-space API and move it into the Linux kernel after people have figured out what that OS should really look like. This is all new stuff.

Looking at the CELL architecture overview [scea.com], though, the Cell doesn't look to me like a desktop replacement. It looks like a video card replacement. Think about it: the biggest piece of closed-source, proprietary hardware in your PC right now is your video card, with its sekrit interfaces and binary-only drivers. We're already starting to see a movement towards more general-purpose use of that hardware with things like nVidia's Cg toolkit [nvidia.com]. The CELL is the logical next step in that direction. You'll have a video card that runs Linux (or, ideally, a video card that acts as just another (heterogeneous) processor in your system).

Re:It has still yet to be explained to me

[mcc (14761)]

I'm asking about how this is implemented. And unless IBM has some kind of insane dynamic recompilation thing going on, the main CPU could not just "divide up the work", since that would entail parallelizing running code across eight processors at once. If the work is already in some way divided I could see the main CPU assigning those divided units to different processors, but even in that case I'd want to know how they implement and make decisions concerning that assignment, and I'm still curious what the SPEs do during a main-cpu context switch.

They have to present the divisions between these SPE processors, or some abstraction which becomes the divisions between them by the time the program is run, visible to the programmer-- since the programmer is the one parallelizing the code. What do these divisions look like to the programmer? Threads? Processes? "Cells"?

Think "Beowulf Cluster" in your living room

[James McP (3700)]

Cell is a multipurpose system. It's main claim to fame is a low-level logic that allows it to farm tasks out to other Cells it connects to dynamically. One Cell is pretty potent and will likely be able to handle the needs of a typical HDTV so IBM hopes every TV, TiVo, and stereo system has a Cell.

The cell system workload sharing system is apparently accessible through the general bus so it can theoretically farm tasks out to any Cell on the same network. So if you've got a WiFi network between your PS4, HDTV, TiVo, Stereo, and cell-powered PDA your video games (or PDA) could take advantage of those other devices' unused clock cycles.

Here's some A to RTF.

http://arstechnica.com/articles/paedia/cpu/cell-1. ars [arstechnica.com]
http://arstechnica.com/articles/paedia/cpu/cell-2. ars [arstechnica.com]

Re:Interesting

[morcheeba (260908)]

... not every processor is documented. In fact, the documentation for a lot of high volume processors isn't available. I'd like PSP docs, or docs on the processor in the ipod, or even the processor in the disposable digital cameras [maushammer.com] or the VMU game system [maushammer.com] -- but none of these docs are publicly available. The hardware manual for the G5 PowerPC [ibm.com] was finally published last November, over a year after apple started selling the hardware [bio-itworld.com].

So, no, the documentation doesn't always get released.

Re:Interesting

[g-san (93038)]

how do i respode to a message?

you are obviously a script.