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."

Re:the real question

[nadamsieee]

From TFA:

The trio is almost done with an application binary interface and language extensions for Cell. A system-level simulator is also nearly complete. Yet to come is a full-fledged Linux implementation for the CPU.

So it sounds like IBM is working on porting Linux to it.

Re:What I wonder...

[bradkittenbrink]

This will blow 3rd party development wide open for the next gen Playstation.

Just like 3rd party development is wide open for the xbox, just because the cpu is publicly documented. Keep dreaming.

Sony must have approved this

[John Seminal]

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:Calling all pawns...

[sampspoon]

nope, they already have cell processor HDTVs on the market

Re:the real question

[KeyShark]

There's already code:

http://ozlabs.org/pipermail/linuxppc64-dev/2005-Ma y/004027.html [ozlabs.org]

Re:All 3 consoles = IBM?

[mrchaotica]

When you say "both next-gen consoles," don't you mean "all three?" The Nintendo Revolution will have an IBM CPU too!

Re:Open HARDWARE movement

[Locke2005]

Yeah, you can simulate it in Verilog (on a high-end Linux server, no less), and get other companies to do your fab for you, but it still costs a minimum of about $100,000 for each chip spin to get actual hardware -- hardly within reach of your average hobbyist. Add to that tens of thousands to license the cores used. I think only a handful of large companies will be designing and building chips for the forseeable future. In fact, if you can't amortize your development costs over about a million chips, it isn't really economically feasible to roll your own.

Re:Open HARDWARE movement

[Vorondil28]

You're absolutely right.

IBM has been doing this for a very, very long time. Ever wonder how the IBM-Compatible PC became so popular, and the Apples didn't? It's simple, IBM kept the AT backplane and CPU architecture open, and Apple didn't. Maintainers of the machines don't have to worry about interoperability, and have a huge list of vendors they can get software and add-on hardware from.

Open hardware is just as important as the software that runs on it.

Re:Interesting

[morcheeba]

... 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:Linux anyone?

[zdzichu]

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

Re:Hardware isn't open

[Blakey Rat]

It's called a "compiler," smartguy.

Re:Open HARDWARE movement

[Anonymous Coward]

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.

Not really that new. Every commercial used CPU out there has really good docs, because they realize that their CPUs are only useful if people have code they can run on them. Hell, I have (or at least had, I may have given them away when I moved) a 3 volume set of printed bound documentation for AMD64, plus a CD with PDFs of same, which AMD sent to me for free just for filling out a form on their website. And I have dozens of PDF and PostScript files for coding for PowerPC, SPARC, Alpha, and several x86 variants. I had been wondering when IBM was going to get around to getting this stuff out there for Cell, since if they want people to use it in workstations, first thing they'll need is a GCC port (I can't see a Windows port to Cell anytime soon, and AIX with XLC just wouldn't be any fun).

The graphics card manufactuers are the exception, not the rule.

Re:What I wonder...

[gamer4Life]

Mods find it "Interesting" that the parent didn't read the article. ;) (Maybe trying to be the first to reply ;))

The article clearly states that Sony is part of this decision.

Re:Sony must have approved this

[apnar]

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.

Interesting comparison. Check out this thread:
http://www.avsforum.com/avs-vb/showthread.php?t=52 7136 [avsforum.com]

It has been broken, at least some of it.

Re:the real question

[Anonymous Coward]

As of 2 years ago, the Nintendo GameCube was the biggest use of PowerPC chips. They told nearly 10x as many GameCubes and Macintoshs.

OpenCores

[jd]

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.

Re:Cell: new desktop processor, or video-card kill

[GROOFY]

That's what Sony thought - but they found it's just not as good as a regular video card is. That's why the PS3 has an nvidia chip instead of a second Cell, as they originally planned.

Re:Sony must have approved this

[IamTheRealMike]

Not quite - it was "broken" in the sense that if you had a valid license you could dump out an unencrypted version that could then be redistributed OK. Obviously:

a) Somebody needs to have bought, cracked and redistributed the media you are interested in

b) That somebody needs to be sure the files aren't watermarked. Presumably if content has been protected it's been paid for, probably by a credit card. So the leaked files could definitely (in theory) be traced back if watermarking was in use. Of course if you can detect it sometimes watermarking can be defeated too but it's a definite risk - are you willing to risk prosection for the priviledge of uploading content?

What that forum also neglects to mention is that Microsoft released an update that kills the crack. The "drmdbg" program can no longer succesfully follow the internal DLL calls inside the DRM software so it cannot extract the key. Because of the way Windows Media works, content can demand that this version of the DRM system is installed before playing so effectively cracks like this can be "closed" quite fast.

So there are two sides to DRM - on one hand, the there's no such thing as "uncrackable" DRM in the absolute sense, on the other hand it's perfectly possible to produce DRM so hard to crack nobody manages it.

Digital satellite DRM is one obvious example of that - modern smartcard security is so advanced that the rollout of new protections like P4 in the states (and I believe they now several generations ahead of what's been cracked) basically eliminated satellite TV piracy. Same is true in the UK - there is no way to pirate Sky TV.

Now, Microsoft has some very smart people working for them, their software is closed source and their DRM is constantly adapting (as the drm2wmv program showed) to close off avenues of attack as they are discovered. They don't have to make it uncrackable, they just have to make it hard and awkward enough to crack that it becomes easier/less risky to buy the content than try and pirate it.

Given that P2P networks are seedy underworlds of trojans, crap content and dodgy downloads, it's not too hard to produce a service that appeals to customers more than attempting to break the DRM does.

Re:Functional Compilers, anyone?

[be-fan]

Part of the advantage of FP is that it can be automatically parallelized much easier than imperative code, and is easier and less error-prone to write than multithreaded imperative code.

What else?

[mnmn]

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?

Re:Sony must have approved this

[LnxAddct]

Here is a walk through kind of thing [blogspot.com]

This was one of many announcements. [msmvps.com]

Most of the news sites that reported it, couldn't confirm it. It was big in japan, but no where else really considering that it isn't just a click and play kind of thing and you already had to have a license to remove the encryption. It's easier to just record your video out and take the minor loss that might be made if you use a cheap cable. But still... programs do exist to pull the drm key out of memory while the video is playing and then another tool exists to remove the DRM using said key. Its a pain in the ass but doable. The exe is no longer on websites (none that I know of)but you might find it on p2p networks. Rumor has it that Micorosft handled the situation quickly and silently with lawyers. Regardless, in the middle of February or so, Microsoft released an update that apparently makes this tool useless and allows content providers to make sure that you have the update installed. Treacherous computing at its best.
Regards,
Steve

Re:What I wonder...

[jericho4.0]

Details are sketchy on the 360's cpu. "3 64 bit Power cores with altivecs and 2 threads each" pretty much captures it. The description sounds pretty mundane, and is probably just like it sounds.

The Cell, OTOH, seems quite novel, and does some very interesting things, even if it does share the Power core. The published benchmarks are very impressive, though admittedly in a narrower domain than what you might be useing your PC for.

Re:What else?

[turgid]

Sort of. Sun is cancelling all UltraSPARC development except in novel niche designs such as Niagara and ROCK. UltraSPARC IV+ will probably be the last Sun-designed conventional UltraSPARC CPU and is intended to keep things limping along until they transition the entire high-end line to Fujitsu SPARC64. Sun will develop Niagara and Rock separately from Fujitsu, who will develop the more conventional SPARC64.

Sun's CPU design team sucks. It's a huge money sink, consistently delivers processors late, under speed and over budget.

For example, last yeat it cancelled the Millenium project, which was the original UltraSPARC V. It was finally going to do such things as out-of-order execution, which every other major RISC has had since the 90s. It was due to come out in 2000, hence the project name.

Sun has, or had, the second largest CPU design team in the world. Why can't it deliver? Why do they insist on using TI to fabricate their chips? Look at what AMD has managed. Why doesn't Sun ask AMD to make UltraSPARC CPUs for it? The Opteron is a RISC internally with an x86 translation layer. Why couldn't Sun ask AMD to make them an Opteron with an UltraSPARC translation layer on it instead?

Sun should have done something about its CPU operation years ago. Maybe it wouldn't be so far behind Opteron, POWER and PowerPC.

Sun has a world-leading server OS in Solaris. If they had a CPU to match, they would be making a huge profit and the share price would be up. The machines would sell themselves.

You have to buy a very large UltraSPARC IV server to beat the performance of a 4-way Opteron. I don't see any future for SPARC. It's in an ever-shrinking niche. Sun had better start work on 16- and 32- way Solaris Opteron boxes and get busy writing a fast SPARC software emulator.