[Dave J]

Since November 30th (more or less) when IBM launched a revolutionary new processor little has been heard from, arguably, the Power client that should be most interested, Apple Computer.

Sony and Toshiba plan on using 4 of these Cells in their Play Station 3. With each Cell theoretically capable of 250 gigaflops, 1 teraflop of potential performance is realized - in a gaming console. This gem will shake the computing industry to its core and may even unseat the mighty x86.

Here is part of what is known about Cell:

Each cell is composed of a processing unit (970 or derivative) plus 8 attached processing units (APU). Each of these APU's have their own memory and are vector processors (graphics, DSP, audio, etc.)

No cache memory for the APU's is used.

Abstract, object-oriented programming is NOT supported. Thus this is a concrete processor (a la Cray's designs). More work for the programmer with payoff in speed.

Fully scalable. 560 Cells would produce the world's fastest supercomputer. 8000 would have a performance potential of 1 petaflop (1 quadrillion floating ops per second = 1,000,000,000,000,000)

Because of the vast quantities IBM will produce at its Fishkill plant and its new plant now under construction and the millions more produced under license by other manufacturers, Cell will be cheap, cheap, cheap.

Adaptable to many OS's and (under emulation) able to run anything.

Many more tidbits are available here:

http://www.blachford.info/computer/Cells/Cell0.html

(There are 5 parts to Blachford's discussion, each more fascinating than the last. Choose Part 4 and scroll to approximately midway to read about Apple's potential vis a vis Cell.)

Question: what is Apple doing or planning to do with Cell?

[thegibbon]

really interesting stuff - thanks for posting that!

[Luca]

I believe this was discussed a few months ago. There's very little chance of Apple adopting the Cell processor.

[Kickaha]

Originally Posted by Dave J Abstract, object-oriented programming is NOT supported. Thus this is a concrete processor (a la Cray's designs).

Whaaaaaaaa.....?

Care to explain how in the world OO programming could not be supported?

A Turing machine is a Turing machine is a Turing machine.

Edit: read the article. Sounds like a chip to me. The abstraction occurs in the compiler, as it *always* has. (Yes, some chips have on-board 'compilers', but those are very very recent.)

OO will be just as supported on this machine as it is on any other. Sheesh.

[Enki]

Originally Posted by Kickaha Whaaaaaaaa.....? A Turing machine is a Turing machine is a Turing machine.

True, but not all Turing Machines are Von Neumann machines. Parallel and vector machines require a different programming paradigm even if they are still Turing Machines at heart.

Haven't had time to read the link yet, but I would stand out on a limb that that is more what the author wanted to say. I agree it would be awful hard if not impossible to come up with something that was incompatible with OO principles. But it might not be too far from hard to throw out the usefulness of many existing library packages if you were interested in REALLY getting your hz worth in a highly parallelized asymmetric chip.

Some hours later: Well Blanchard is quite sloppy in his facts and writing style. The Commodore Amiga???

His attempt to proclaim the death of OO programming was related to his proclamation on the death of hardware abstraction. But he never made a case for programmer abstraction going away, just the end result code. And if those restrictions are anywhere near true the last 30 years of programming productivity advances just got tossed out the window. We might not ever see truly complex Cell software because it will be just too hard to write effectively without abstraction. Since that whole situation is pretty much unlikely and his other guesstimates not being take-to-the-bankable I would call the article mildly interesting and thought provoking, but not definitive by ANY stretch of the imagination.

[mama's left eye]

Does sony have big part in it's developement? If they do then 250 gigaflops could very well mean 250 kilaflops...

[Kickaha]

Originally Posted by Enki True, but not all Turing Machines are Von Neumann machines. Parallel and vector machines require a different programming paradigm even if they are still Turing Machines at heart. Haven't had time to read the link yet, but I would stand out on a limb that that is more what the author wanted to say. I agree it would be awful hard if not impossible to come up with something that was incompatible with OO principles. But it might not be too far from hard to throw out the usefulness of many existing library packages if you were interested in REALLY getting your hz worth in a highly parallelized asymmetric chip.

Here's the quote:

The Cell architecture goes against the grain in many areas but in one area it has gone in the complete opposite direction to the rest of the technology industry. Operating systems started as a rudimentary way for programs to talk to hardware without developers having the to write their own drivers every time. As time went on operating systems have evolved and taking on a wide variety of complex tasks, one way it has done this is by abstracting more and more away from the hardware. Object oriented programming goes further and abstracts individual parts of programs away from each other. This has evolved into Java like technologies which provide their own environment thus abstracting the application away from the individual operating system. Web technologies do the same thing, the platform which is serving you with this page is completely irrelevant, as is the platform viewing it. When writing this I did not have to make a Windows or Mac specific version of the HTML, the underlying hardware, OSs and web browsers are completely abstracted away. Even hardware manufacturers have taken to abstraction, the Transmeta line of CPUs are sold as x86 CPUs but in reality they are not. They provide an abstraction in software which hides the inner details of the CPU which is not only not x86 but a completely different architecture. This is not unique to Transmeta or even x86, the internal architecture of most modern CPUs is very different from their programming model. If there is a law in computing, Abstraction is it, it is an essential piece of today's computing technology, much of what we do would not be possible without it. Cell however, has abandoned it. The programming model for the Cell will be concrete, when you program an APU you will be programming what is in the APU itself, not some abstraction. You will be "hitting the hardware" so to speak.

All that is really saying is that you write to silicon, not to an OS, not to a JIT, not to a runtime, not to a pipeline allocator. Which is what basically every compiler on the planet does in the final analysis. It's just an open and direct assembly language, is all that's really saying. Should make targeting the chip easy for a compiler, but harder to optimize in some ways. Now, of course, if they stuck to a nice simple RISC set, that task should be easier.

In the end though, nothing to see here citizens, this is just getting back to basics.

[Enki]

Agreed. I was reading the article and composing my above edit while you were responding. If I believed he was balls on I would think Cell would have an incredibly difficult time getting out of the easily parallelized niche due to that back to basics. It's still wait and see though.

[MarkTime]

I want to explain the puzzling Amiga reference.

The Amiga community, like this Mac community, is still alive and kicking.
You can view it at www.amiga.org, www.ann.lu, and www.amigaworld.net

While this community is only a remnant of the larger one that existed when Commodore Amigas were still relevant...you have to understand the culture of this small community, to understand why someone thought pulling an Amiga reference might be understood...in that one small segment of the planet, the reference was understood.

Well, except even we criticized him for that remark....

Let me say this, about the Amiga community, what is left of it is some highly intelligent people and some pretenders....even though we mostly don't use actual Amiga's anymore, the community is still a wonderful place that hearkens back to the early days of the community.

You know, that old community, when everyone involved were technology pioneeres, before the AOL induced masses came in and flooded out all the new reasoning with their talk of Bennifer and their reactionary politics, like you see in the reaction to ThinkSecret's lawsuit.

Come by, visit for a while...as for me, I'm glad to be here...this is the first forum site in the Mac community, that I've found that was interesting all.

I agree with your assessment of Nicholas' article on Cell...it's really just a word, talking about concepts that are as old as computing itself, but...can't blame a kid for trying. Amiga kids are talking about Cell, and at least trying to push the envelope...when I jumped into this community, in the middle of the ThinkSecret controversy, it seemed to me Mac kids were all about trying to explain why free speach can never really be unlimited, and one particulary intelligent child even said 'you should really listen to what big corporations tell you to do.'...wow, a community that thinks fascism is cutting edge.

but having found this forum, I will rethink the Mac community....with the great hardware and cutting edge OS...I've been thinking of dumping the old Amiga community, for something...a little more new.

[Dave J]

I for one lamented the passing of the Amiga, first hearing about it when it was the Lorraine as Blanchard states, then rejoicing when I read that CBM would pick it up thinking - erroneously - that Commodore would do a respectable marketing job on it (a la the C64, C128).

I rejoiced too soon. Commodore had one asset: Jack Tramiel. When he was voted out of the top slot, Commodore began to make one misstep after another. (Jack eventually went to Atari where he made a Herculean, but eventually fruitless, effort.)

[oldmacfan]

I spent the better part of yesterday reading the interpretation and will be very excited to read more on this technology when it is made available.

As for whether or not Apple will some how use the Cell in a Mac, well you know Apple would have never ever ever joined on as a named partner. It would just create to many problems for them. I do believe that Apple will be looking very closely at the Cell, but this will never get publicly acknowledged unless they release a product with it.

Remember Keynote and Sony President? Remember the odd speculation that has occurred since then regarding the two companies?

[Dave J]

Originally Posted by oldmacfan As for whether or not Apple will some how use the Cell in a Mac, well you know Apple would have never ever ever joined on as a named partner. It would just create to many problems for them. I do believe that Apple will be looking very closely at the Cell, but this will never get publicly acknowledged unless they release a product with it. Remember Keynote and Sony President? Remember the odd speculation that has occurred since then regarding the two companies?

I agree and think it is premature to dismiss Cell as a strictly auxiliary "game-station-solution". In the end it all comes down to the money. IBM and its 16 Cell partners are plunking down billions to build the 2 65 nm factories. Clearly for IBM at least, Cell is a major player.

AFA Apple's involvement - purpose of this thread - they've got to be making the comparison between the 250 gigaflops of 1 Cell and the paltry 17 of the dual 2.5 G5.

[oldmacfan]

From the little I do understand about the underpinnings of OS X there would have to be a major rewrite of code to optimize it for a cell.

In another thread, I mentioned the idea of using it as a Co-processor of sorts. You could have computers that are hybrids while code rewrite is being worked on.

Start with a little recoding to take some advantage of the technology and then as more and more new code comes online (which takes time) you patiently move in that direction.

I would speculate that if OS X (10.4 or 5) starts to move towards more and more threading of the OS and applications, and with the industry hype of multi-core chips, we could see Apple taking part in Cell.

The part of the article that I liked was the idea that Sony could make a version of PS3 that would be an add-in card for your computer. That I might consider buying. It would truly give my dust collector (PC) something to do.

[Dave J]

Originally Posted by oldmacfan In another thread, I mentioned the idea of using it as a Co-processor of sorts. You could have computers that are hybrids while code rewrite is being worked on.

(And I suggested a PCI-X.) I like your suggestion better though. With all that vector processing on tap who would need a dedicated GPU? And GPU's are already being asked to do something other than push pixels around.

[Jim S.]

Let's see...

1. Sony has a major investment in cell processors
2. Sony would like to build computers with cell but with what OS?
3. Sony owns a lot of movies
4. Apple wants to sell movies and video pods
5. Apple wants market share
6. Apple admires Sony
7. Cell processors are thought to be "the future of computing"
8. Apple would like to blow away the Intel world

Can anyone make any sense of this?

[oldmacfan]

Originally Posted by Jim S. Let's see... 1. Sony has a major investment in cell processors 2. Sony would like to build computers with cell but with what OS? 3. Sony owns a lot of movies 4. Apple wants to sell movies and video pods 5. Apple wants market share 6. Apple admires Sony

6.5) Sony admires Apple and once again wants to be looked upon in those eyes.

Originally Posted by Jim S. 7. Cell processors are thought to be "the future of computing" 8. Apple would like to blow away the Intel world Can anyone make any sense of this?

[MarkTime]

Wow, you all certainly get into the fantasy side of computing.

Cell is nothing but some words, even though I would credit Nicholas for being an ambitious nerd, he doesn't really have any particular insight into Cell.

[oldmacfan]

Originally Posted by MarkTime Wow, you all certainly get into the fantasy side of computing. Cell is nothing but some words, even though I would credit Nicholas for being an ambitious nerd, he doesn't really have any particular insight into Cell.

Very true, but did you read the forum title...

Yes, Nicholas is an ambitious nerd, he read the patent application and all the other documentation that he could come up with and then combined this with his knowledge set and wrote a speculative piece. This is something that happens every day on TV news and then the supposed experts they drag on camera usually haven't any real inside information and might not even be versed on the situation.

As for the fantasy side of computing.

Here’s to the crazy ones.

The misfits.

The rebels.

The troublemakers.

The round pegs in the square holes.

The ones who see things differently.

They’re not fond of rules.

And they have no respect for the status quo.

You can praise them, disagree with them, quote them,

disbelieve them, glorify or vilify them.

About the only thing you can’t do is ignore them.

Because they change things.

They invent. They imagine. They heal.

They explore. They create. They inspire.

They push the human race forward.

Maybe they have to be crazy.

How else can you stare at an empty canvas and see a work of art?
Or sit in silence and hear a song that’s never been written?
Or gaze at a red planet and see a laboratory on wheels?

We make tools for these kinds of people.

While some see them as the crazy ones,
we see genius.

Because the people who are crazy enough to think
they can change the world, are the ones who do.

Now why did you make me do that?

[Philth]

No one will have a choice if fantasy becomes reality.

[TednDi]

Wait...


what is going on with Benifer?

i N E E D to Know!!

[oldmacfan]

Now this is a perfectly good Cell thread...

[DrGruv]

methinks people are a little to uh... nevermind (just joking)

[BenRoethig]

Well, they're supposed to have some kind of press conference about the cell on Monday. I'll very interested to find out if its all its rumored to be.

[Dave J]

Originally Posted by BenRoethig Well, they're supposed to have some kind of press conference about the cell on Monday. I'll very interested to find out if its all its rumored to be.

Yes. It's the ISSCC. (Much more than a press conference though.) Worth following. Here's the link:
http://www.isscc.org/isscc/

[Chinney]

Latest link. Nothing new, really, but makes the previous information more official.

http://news.zdnet.co.uk/hardware/chi...9186973,00.htm

I am still completely unclear about what this means for the personal computer market (Intel, MS, Apple, Sony, IBM…anyone) two-three years down the road.

[Dave J]

News is trickling in from the ISSCC. Full Doc available for download here:

http://www.scee.presscentre.com/imag...etailsID=25555

Selected quotes:

"Cell is OS neutral - supporting multiple operating systems simultaneously"

"Cell's breakthrough multi-core architecture and ultra high-speed communications capabilities deliver vastly improved, real-time response for entertainment and rich media applications, in many cases 10 times the performance of the latest PC processors."

"Cell is a breakthrough architectural design -- featuring eight synergistic processors and top clock speeds of greater than 4 GHz (as measured during initial hardware testing)"

"We are proud that Cell, a revolutionary microprocessor with a brand new architecture that leapfrogs the performance of existing processors, has been created through a perfect synergy of IBM, Sony Group and Toshiba's capabilities and talented resources, "said Masashi Muromachi, corporate vice president of Toshiba Corporation and president & CEO of Toshiba's Semiconductor Company. "We are confident that Cell will provide major momentum for the progress of digital convergence, as a core device sustaining a whole spectrum of advanced information-rich broadband applications, from consumer electronics, home entertainment through various industrial systems."

[Stephen Wright]

I think some of you have jumped to teh wrong conclusion about Cell, yea its based on a G4(ish) CPU but it has been heavily modified, however a Mac is a media machine and as much can never really have enough grunt as such.
Now the other partner in the Ps3 project is Nvidia, so think of it this way, you will have your G5's clocked to whatever, a VERY fast bus (hypertransport?) connecting the CPU - GPU/CELL and as much as possible your drivers / os can be tuned to pushing all media processing onto the GPU/CELL chip.

[Bofferma]

The next generation consoles will be a huge threat to Apple. Sony has no reason to cede the "video on demand" market to some Apple iMovie Store. The PS3 and XBox2 will have the horsepower and connectivity to handle all your hi-def audio and video needs--and coupled with a HDTV monitor they have the potential to really hurt the personal computer market. Steve might be a happening dude with good hollywood contacts--that doesn't mean MS or Sony will cut him a slice.

If Apple ceases to be percieved as that "hip media company" they will suffer mightily. The iPod and iTunes music store have given them an edge for the past few years, but if they don't move quick they will return to being that flailing computer maker (albeit with a nice OS and industrial design)--rich school districts and snobby photoshop users need only apply. Keep an eye on your Apple stock... it might be time to sell.

They have a very small time window in which to carve out a niche before the consoles move in. Too bad IBM couldn't provide them with that quick low-power chip they really needed for the mini.

Hmm... do I sense a conspiracy?

[DrGruv]

APPLE IS THE NEXT MICROSOFT (buy stock now)

can't you hear it?

Well the chip is great, but there's no software to run on it... *insert apple here

[Bofferma]

Are you implying that the PS3 will be "Powered by MacOS"? Why?