Sign in with
Sign up | Sign in
Your question

The Cell Processor

Tags:
  • CPUs
  • Sony
  • Processors
Last response: in CPUs
Share
March 11, 2006 3:55:36 PM

Will all this talk of Intel and AMD, I thought I would take the liberty to speak of a lesser-known upcoming processor from Sony. Despite me being an XBOX person, I am interested in this. It supposedly has seven cores running 3.2 GHz each, and one more of the same reserved for "redundancy." Sony announced some time ago that this processor which is also schedualed to be placed in the PS3 (meh), will also be released to various computer manufacturers (IBM included, they helped develope it). If you wish for some kind of a pole, list the "choices" and Ill post a poll. Voice your opinions here.

More about : cell processor

March 11, 2006 4:13:22 PM

Quote:
Will all this talk of Intel and AMD, I thought I would take the liberty to speak of a lesser-known upcoming processor from Sony. Despite me being an XBOX person, I am interested in this. It supposedly has seven cores running 3.2 GHz each, and one more of the same reserved for "redundancy." Sony announced some time ago that this processor which is also schedualed to be placed in the PS3 (meh), will also be released to various computer manufacturers (IBM included, they helped develope it). If you wish for some kind of a pole, list the "choices" and Ill post a poll. Voice your opinions here.


The Cell processor is really junk, with the crap coherency between the SPE's and the lack of enough cache, the Cell really isn't as powerful as they chock it up to be.

~~Mad Mod Mike, pimpin' the world 1 rig at a time
March 11, 2006 4:20:12 PM

yeah, this processor's gonna be great, but think about a) the heat b) the size
(i highly doubt they'll fit it on one of our normal chips (in fact i can almost guarantee they won't)) c) what proprietary (spelling?) socket are they going to come up with d) how much will it cost?

i think it it will generate a lot of heat, i believe easily twice the prescotts (spread about of course due to larger surphace area) and i think they are probably going to charge a lot for their own socket and design (motherboards too? and RAM? (don't forget their working with Rambus....))

Ara

then again, it could be a big flop
Related resources
March 11, 2006 4:22:42 PM

The Cell processor is garbage.
March 11, 2006 4:23:04 PM

Quote:
Will all this talk of Intel and AMD, I thought I would take the liberty to speak of a lesser-known upcoming processor from Sony. Despite me being an XBOX person, I am interested in this. It supposedly has seven cores running 3.2 GHz each, and one more of the same reserved for "redundancy." Sony announced some time ago that this processor which is also schedualed to be placed in the PS3 (meh), will also be released to various computer manufacturers (IBM included, they helped develope it). If you wish for some kind of a pole, list the "choices" and Ill post a poll. Voice your opinions here.


The Cell processor is really junk, with the crap coherency between the SPE's and the lack of enough cache, the Cell really isn't as powerful as they chock it up to be.

~~Mad Mod Mike, pimpin' the world 1 rig at a timeI never thought so, but who knows, it could surprise us :roll: . Not like I will get it anyway. It makes me mad to, every time I go to bestbuy or someplace, they had X360's THE DAY BEFORE I CAME. It angers me so.
March 11, 2006 4:26:15 PM

Quote:
Will all this talk of Intel and AMD, I thought I would take the liberty to speak of a lesser-known upcoming processor from Sony. Despite me being an XBOX person, I am interested in this. It supposedly has seven cores running 3.2 GHz each, and one more of the same reserved for "redundancy." Sony announced some time ago that this processor which is also schedualed to be placed in the PS3 (meh), will also be released to various computer manufacturers (IBM included, they helped develope it). If you wish for some kind of a pole, list the "choices" and Ill post a poll. Voice your opinions here.


The Cell processor is really junk, with the crap coherency between the SPE's and the lack of enough cache, the Cell really isn't as powerful as they chock it up to be.

~~Mad Mod Mike, pimpin' the world 1 rig at a timeI never thought so, but who knows, it could surprise us :roll: . Not like I will get it anyway. It makes me mad to, every time I go to bestbuy or someplace, they had X360's THE DAY BEFORE I CAME. It angers me so.

Game Developers as well have stated (according to some websites) that the Cell Processor is Heavily overrated and underpowered.

~~Mad Mod Mike, pimpin' the world 1 rig at a time
March 11, 2006 4:27:27 PM

Quote:
yeah, this processor's gonna be great, but think about a) the heat b) the size
(i highly doubt they'll fit it on one of our normal chips (in fact i can almost guarantee they won't)) c) what proprietary (spelling?) socket are they going to come up with d) how much will it cost?

i think it it will generate a lot of heat, i believe easily twice the prescotts (spread about of course due to larger surphace area) and i think they are probably going to charge a lot for their own socket and design (motherboards too? and RAM? (don't forget their working with Rambus....))

Ara

then again, it could be a big flop
I think they were supposed to use a special z-ram or something. And it is actually supposed to run pretty cool compared to today's processors. Im using my laptop because my main machine is down for a stupid reason. Can you guess? My keyboard went out. Meh.
March 11, 2006 4:47:52 PM

I can't wait to purchase one; I'll put it in my basement so it can collect dust with the rest of my consoles.
March 11, 2006 5:29:37 PM

What about if you modified one with Linux, then set it up to run BOINC/FAH? Doing 7/8 WU at once wouldnt be bad.
March 11, 2006 6:33:57 PM

Quote:
Game Developers as well have stated (according to some websites) that the Cell Processor is Heavily overrated and underpowered.

~~Mad Mod Mike, pimpin' the world 1 rig at a time


I don't think Cell will be that great of a game processor, though it does have potential once programmers learn how to take advantage of it.

I think it will really shine in "Big Iron" situations such as datamining, database apps, large scale data processing (weather, geologic data, etc). I saw an article recently discussing how well the Cell Blade servers work for those types of apps, but I can't find it now.

BTW, for those who don't know, Cell is not, nor is it meant to be x86 compatible, so you won't be seeing it on your desktop.
March 11, 2006 6:51:45 PM

I agree with your assessment. Cell processors can be thought of as "application specific processors" ... they "do" one thing, or group of similar things, extremely well because they are dedicated to that one "job". So, ganging a bunch of 'em together for data mining, or the government's scanning of every single electronic transmission on the whole planet is both logical, and factual reality :D 
March 11, 2006 6:54:15 PM

Well, in that respect, they been doing it for 20 years, so now they just have a better way. :twisted:
March 11, 2006 7:08:09 PM

Amen! And, "better way" = greater control of the unwashed masses, doesn't it? :D 
March 11, 2006 11:44:08 PM

Quote:
Will all this talk of Intel and AMD, I thought I would take the liberty to speak of a lesser-known upcoming processor from Sony. Despite me being an XBOX person, I am interested in this. It supposedly has seven cores running 3.2 GHz each, and one more of the same reserved for "redundancy." Sony announced some time ago that this processor which is also schedualed to be placed in the PS3 (meh), will also be released to various computer manufacturers (IBM included, they helped develope it). If you wish for some kind of a pole, list the "choices" and Ill post a poll. Voice your opinions here.


THE CELL IS A PIECE OF **** AND YOU ARE A NEWB!! STFU!! BOW DOWN TO MADMODMIKE U !!

How many Cell processors have you used? None? Good, then shut up, that has no place here one way or the other.
March 12, 2006 2:00:16 AM

Quote:
Will all this talk of Intel and AMD, I thought I would take the liberty to speak of a lesser-known upcoming processor from Sony. Despite me being an XBOX person, I am interested in this. It supposedly has seven cores running 3.2 GHz each, and one more of the same reserved for "redundancy." Sony announced some time ago that this processor which is also schedualed to be placed in the PS3 (meh), will also be released to various computer manufacturers (IBM included, they helped develope it). If you wish for some kind of a pole, list the "choices" and Ill post a poll. Voice your opinions here.


THE CELL IS A PIECE OF **** AND YOU ARE A NEWB!! STFU!! BOW DOWN TO MADMODMIKE U !!

How many Cell processors have you used? None? Good, then shut up, that has no place here one way or the other. I never saw that post, must have got deleted. Oh and by the way "noob" is the correct term in your matter of speaking. I never said I liked the Cell, just said I was interested and could find sparce information because Sony is tight-lipped. I bow to no person, I respect anyone who doesn't completely make a fool of themselves, **cough, you, cough, piddy(joke)** but that is as far as it goes. Look now I'm getting all worked up over a senseless statement.
March 12, 2006 2:42:17 AM

Actually... I remember it originally being "newb".
March 12, 2006 2:52:41 AM

Quote:
Actually... I remember it originally being "newb".
Weird shi7 gets invented in online gaming. Newb started as a short term for being newbie, I don't know whats so hard about typing that extra two letters that make you seem all that much smarter.
March 12, 2006 4:43:03 AM

I think the main reason people consider the cell processor to be a pile of crap is it can only carry out floating point calculations really well but nothing else. In other words, it could simulate fragments of a building bouncing around and tons of shit like that, but would struggle to encode a video. so it fits perfectly into the role of a gaming system due to the extreme visual effects and crazy ass graphics of upcoming games.

The main reason developers hate the idea of this processor is the unique and difficult way they will have to program games for this processor due to the multiple threads they will have to produce. for instance. They would have to have one thread or execution to be performed by one core of the cell processor and another thread to execute be executed by another. usually this can all be done by one processor and each execution is done every second clock etc. Why am i explaining this, you guys probably already know this crap lol.

point being, the Cell is only good for games from what i've read.
March 12, 2006 4:50:53 AM

The main problems with cell are:

1. No OOOe.
2. Limited local storage for the SPEs and they're not cache.
3. Asymetric design.
4. Long pipeline.
5. Weak branch predicton.

Its pretty solid, just not for the PC.
March 12, 2006 4:52:57 AM

Quote:
Game Developers as well have stated (according to some websites) that the Cell Processor is Heavily overrated and underpowered.

~~Mad Mod Mike, pimpin' the world 1 rig at a time


I don't think Cell will be that great of a game processor, though it does have potential once programmers learn how to take advantage of it.

I think it will really shine in "Big Iron" situations such as datamining, database apps, large scale data processing (weather, geologic data, etc). I saw an article recently discussing how well the Cell Blade servers work for those types of apps, but I can't find it now.

BTW, for those who don't know, Cell is not, nor is it meant to be x86 compatible, so you won't be seeing it on your desktop.

Its not that programmer "KNOW" how to take advantage of it, LOL.
Its the Physics systems thats hard to REprogram, yes you would have to reprogram the physic engines base to take advantage of multicore processer, thats why it took physic engine companies so long to actually able to release a stable one out, the Second most intensive task for the CPU in a game is the Physics, AI split to second core?, thats stupid, for most games today, as most AI in games now(like F.E.A.R.) are not very good at thinking, they just react to what you did to them, now thats stupid.
March 12, 2006 5:24:51 AM

The wikipedia entry is a decent read. The thing definately isn't a desktop processor, but it might be a decent server chip. It basically sounds like a GPU to me, with the but I don't know much, so I could be wrong. The XDR sounds nice though...
March 12, 2006 6:20:35 AM

I think SUN’s T1 (Niagara) showed that a ton of simple cores can produce tremendous results. They made their processor to excel at integer buts that’s their niche.

Although the cell is much different, it follows the same recipe. I think it may be even more balanced since each “cell” can push FP as well as INT.
March 12, 2006 7:23:32 AM

Quote:
Game Developers as well have stated (according to some websites) that the Cell Processor is Heavily overrated and underpowered.

~~Mad Mod Mike, pimpin' the world 1 rig at a time


I don't think Cell will be that great of a game processor, though it does have potential once programmers learn how to take advantage of it.

I think it will really shine in "Big Iron" situations such as datamining, database apps, large scale data processing (weather, geologic data, etc). I saw an article recently discussing how well the Cell Blade servers work for those types of apps, but I can't find it now.

BTW, for those who don't know, Cell is not, nor is it meant to be x86 compatible, so you won't be seeing it on your desktop.
I actually have faith in that little chip.
8 Cores? Jesus H Christ...can anybody say long programming hours?
For supercomputing/business needs/servers, Cell could be extremely successful. For gaming...Meh.
We really could see Cell's appear in supercomputing applications though...bloody shame it isn't x86 compatible so we can't compare it head-to-head with Pentium D's and X2's. However, there is good news: supposedly PS3 will be shipping with Linux ready to be used =D
March 12, 2006 7:24:35 AM

Quote:
What about if you modified one with Linux, then set it up to run BOINC/FAH? Doing 7/8 WU at once wouldnt be bad.

PS3 may ship with Linux? Not sure, may want to look that one up.
a b à CPUs
March 12, 2006 7:36:47 AM

Quote:
Will all this talk of Intel and AMD, I thought I would take the liberty to speak of a lesser-known upcoming processor from Sony. Despite me being an XBOX person, I am interested in this. It supposedly has seven cores running 3.2 GHz each, and one more of the same reserved for "redundancy." Sony announced some time ago that this processor which is also schedualed to be placed in the PS3 (meh), will also be released to various computer manufacturers (IBM included, they helped develope it). If you wish for some kind of a pole, list the "choices" and Ill post a poll. Voice your opinions here.


If its like IBMs last great adventure, they will make the concept, and Intel and AMD will take over like x86.
March 12, 2006 9:15:20 AM

(Question) How many 6th/7th gen x86 cores would fit in the space of a 640 million (640 x 10^6) transistor processor (65nm can do this) if a suitable amount of cache is used (and said cache was using Z-RAMs).

- Note that said cores include OoO / Register renaming, etc, etc

The clock speeds would still be in the 2 to 3 GHz range.

So I decided to do some research....

(Answer) 6 x 6/7th generation x86/x64 processor cores, with 'only' 6 MB shared L2 cache, would fit within 630 million transistors, leaving space for 10 million more (extra logic, etc).

TDP for such a processor would be be 75 - 90 watts, (depending on load), which can be cooled by todays 'standard' heatsink + fan combinations (thanks to the PreScott we have some rather advanced cooling technology available :p , ironic no ?).

Power usage would be up to 115 watts (give or take), which is 'acceptable' aswell, and likely lower assuming 4 of the 6 cores are in power saving modes most the time, leaving the 6 MB shared L2 cache to only 2 or 3 of the cores in 'actual usage' patterns.... at least in most software.

The processor would require an external (thin but high speed, likely 64 to 128 bit 'thin') link northbridge / MCH as it would be to complex if the memory controller was integrated within it, and it would require somewhere in the order of between 12.8 and 25.6 GB/sec of peak memory throughput (likely 'wider' at 256 - 512 bit, in total anyway) to keep the cores well fed with instructions and data.

eg: Sort of like how the typical current Intel FSB is 800 MHz (QDR) x 64 bits wide for 6.4 GB/sec, but they use 400 MHz (DDR) x 128 bit wide (dual channel) memory to feed it, up to, the same 6.4 GB/sec.

This could be done using todays 65nm technology, ironically XDR RAMBUS would be the best choice for it aswell (lowest pin count to high data transfer rate ratio than any other solution)..... however I don't see Intel going back to them for 'help'. (Pin counts would become an issue to get enough sustained memory throughput).

It is possible 2 x CPU socket systems using 2 x ccNUMA nodes may become the norm just to aggregate and 'share' the total available memory throughput to keep performance scaling.

Because of manufacturer control it would be unlikely to happen for some time, and Intel tend to burn space on large L2 caches IMHO when an increase in L1 (data and instruction) cache sizes would be a more effective way (transistor wise) to keep performance up. As 6 cores + 12 MB cache would require around 840 million transistors it is not likley to happen with Intel until 45nm manufacturing begins.

eg: Intel would pair 12 MB with 6 cores most likely, maybe even more, and then keep the interface to memory slower, while keeping 'overall complexity' down to reduce mainboard costs. (Frankly I'd rather pay up to $500 for a processor, and up to $500 for a mainboard, excluding RAM, and 'typically' expect to pay $750 for video card).

eg2: A 1 billion (10^9) transistor processor could house 40 Celeron (P6) processor cores, including the ~ 5 MB cache (ideally shared) L2 cache, and overheads. Such as logic to control the large number of cores and a shared interface to memory for them all. (Now that is cool, as they'd be clocked at over 3 GHz using a 52nm, or more likely 45nm, manufacturing process). Go ahead crunch the numbers yourself.... it sounds pretty cool to me. :) 

IBM Cell, just like Intel IA-64 (Itanium), requires a very good compiler, that is where many of the advancements are made, and the point is to make the Out of Order / Register Renaming units on the processors take up less silicon space, so more can be dedicated to combinations of the FPU (Floating Point Unit), cache(s), and other processor cores.

Eventually we'd hit a point where the Out of Order / Register Renaming units grew so large they'd need half the processor space to keep scaling performance, when (or if) we hit that point without moving to another architecture gamers will not be a happy market.

CPUs (General & Central processing units) are going to become more like GPU / VPUs are today... we don't bash ATI or nVidia for not having advanced, large, and complex Out of Order execution units in their GPU / VPUs, and physics calculations on 1,000's of objects at once do not require OoO feaures to make effective use of 'silicon real-estate' within a processor.

==================

Cell is just a cut down (cheaper to produce and scaled) varient of the above.... (that still has good performance) as you don't require Out of Order execution when you can execute multiple threads at once in well coded and compiled software, especially if each core can issue 3 - 4 instructions per clock cycle.

Cell has applications in gaming, but it might take awhile before people migrate to it for gaming.... it also has excellent video encoding and decoding performance for a 'single chip' solution. Bear in mind gaming consoles are getting closer and closer to Desktop PCs in functionality (eg: surf the net, while watching a DVD, BluRay, HD-DVD, etc. How long would it be until consoles can encode video faster, and better, than existing PCs ?).

The compilers job moves to 'making code that avoids pipeline stalls at all costs', which isn't really that much of a jump for compilers anyway, and less on making (streams of) code that can be executed out of order.

It is a far more elegant solution and will result in performance scaling far better than it has since 2002.

Really since 2002 has CPU performance, excluding the 2 x 4 issue core Intel Conroe, really scaled that well compared to previous 3 to 4 year periods ?
March 12, 2006 10:11:12 AM

Or,... in more laymens terms:

If multiple branches and/or possibilities of the code can be run in parallel, (similar to the Quantum computing concepts in a way), aswell as multiple loops of code running 'mostly' independant of each other, do we still require advanced branch prediction and out of order execution to go with it ?

The answer is no, we'll only require more basic branch prediction and very basic out of order execution with 'smarter' code, especially when upwards of 128 instuctions (over 32 threads) can be fetched and processed per clock cycle.

It would still require a smart and fairly complex scheduler, to feed entire branches of code to multiple processor cores at once, and just ignore any wasted branches that have been executed. (Wasteful perhaps, but still lightyears ahead of what we are doing now).

But that is where the compiler needs to make significant jumps forward (they have not changed much since x86's inception, ideally we should be moving to a new architecure, not x64 extentions either, by around 2008 to 2012).

Intel want this architecture to be IA-64, or a combiniation of x86/x64 and IA-64 within a single box.
!