IBM's Plans For the Cell Processor 124
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."
With more memory per CPU, it might not suck (Score:5, Interesting)
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.
IBM - proven in this market space (Score:2, Interesting)
They could actually try to sell the Cell (Score:5, Interesting)
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.
Re:Whats a Future Power Road Map? (Score:5, Interesting)
You've really missed hearing about Cell?
It's a new processor architecture, IBM and Sony (and possibly others) had a hand in it. Effectively two "Power" cores and a bunch of vector processing units. It's supposed to be very very good for vector operations. For a while (a few years back now) the world's most powerful supercomputer was a machine composed of nodes containing two cell processors and an Opteron each.
It's different to other parallelisation strategies as the vector units (SPU/SPEs) allow you to parallelise stuff at an operation level, unlike just stuffing more cores into the box which is the intel/PC strategy. For games and graphics this it thought to be good, hence its inclusion in the playstation 3. It's also supposed to be good for scientific computing.
I guess you could think of it as somewhere between a CPU and a GPU, or a hybrid of the two approaches.
what would be cool (Score:3, Interesting)
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.
Comment removed (Score:5, Interesting)
Laughable Drivel (Score:2, Interesting)
"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 architecture as painful as the Cell."
"tough to program."
It's like you tried to parrot every Beyond3d x86 fanboy talking point you could remember.
Re:They could actually try to sell the Cell (Score:5, Interesting)
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."
Cell Is Being Used Exactly How It Was Designed For (Score:3, Interesting)
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 major areas are running full speed in parallel. Split memory architecture lets each area of the machine run at full speed without interfering with the rest of the system.
Kutagari and IBM did a masterful job. It was an obvious choice to build off the model of the most sucessful console architecture in history and the one all console developers had intimate knowledge of, the 145 million selling PS2.
Re:With more memory per CPU, it might not suck (Score:3, Interesting)
The SPE's aren't full CPU's, they're essentially enhanced versions of the PS2's VU's.
Yep, stream data, just like on the PS2.
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), small pans (VU/SPE RAM)" http://arstechnica.com/old/content/2000/04/ps2vspc.ars/1 [arstechnica.com]
The Cell isn't MIPS, it's PPC, the PS2 (and PS1) were the MIPS machines. The SPE's are supposed to handle things like audio and networking, that's their job. Apparently you can also do things like assign a SPE to do things like very fast bzip decompression.
Re:Cell Is Being Used Exactly How It Was Designed (Score:3, Interesting)
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]