bit-tech.net

Intel builds 80-core prototype

Intel builds 80-core prototype

The Tera-Scale Teraflop Prototype chip looks promising for the future - energy efficiency is a major focal point.

Intel's research team has managed to successfully produce a prototype 80-core Tera-Scale processor that uses less energy than the company's current flagship Core 2 Extreme QX6700 quad-core processor.

The prototype was built so that the chip giant's researchers could investigate the best way to make such a large number of processing cores communicate with each other. This was in addition to researching new architectural techniques and core designs.

The chip, dubbed the Tera-Scale Teraflop Prototype, is just for research purposes and lacks a lot of necessary functionality at the moment. However, R&D Technology Strategist Manny Vara said that the company will be able to produce 80-core chips en masse in five to eight years.

Currently, the prototype chip consumes less than 100W of power, which is less than the 130W consumed by the quad-core QX6700. Of course, the prototype currently lacks some key functionality, which could potentially throw the power consumption characteristics out of proportion, but it's an impressive feat nonetheless.

Vara added that although there are many more cores on the Tera-Scale prototype, they're a different type of core than the ones used in today's microprocessors. "The new ones will be much simpler. You break the core's tasks into pieces and each task can be assigned to a core. Even if the cores are simpler and slower, you have a lot more of them so you have more performance."

Today's microprocessor cores are very flexible, while Intel believes that tomorrow's microprocessor cores will be much more specialised, but of course, there will be many more of these simpler cores. AMD's Fusion project appears to be going down the route of scaling what we've already got, while Intel is moving towards what would be a more flexible approach to energy efficiency.

Before you get too excited though, this is all on paper at the moment; the real war of the cores won't be decided until both companies have released their respective massively multi-core processing architectures in a few years time.

30 Comments

Discuss in the forums Reply
Leitchy 18th January 2007, 12:32 Quote
40 cores dedicated to Windows Vista and 40 cores for other apps.
r4tch3t 18th January 2007, 12:44 Quote
Quote:
Originally Posted by Leitchy
40 cores dedicated to Windows Vista and 40 cores for other apps.
You really don't like MS do you?

I think that this may be a step in the right direction, but it could just turn out to be the next GHz race.
DeX 18th January 2007, 12:45 Quote
How measly will 2 cores appear in a few years' time? The race will be one for the first kilo-core processor. :p
DXR_13KE 18th January 2007, 12:49 Quote
maybe my future cpu... let the technology mature enough :D
Leitchy 18th January 2007, 13:15 Quote
Quote:
Originally Posted by r4tch3t
You really don't like MS do you?

I think that this may be a step in the right direction, but it could just turn out to be the next GHz race.

Love Microsoft personally, Since I use them everyday in my Job and at Home, I'm just a bit biffed at my Vista installation last night blue screening, then hogging all my resources when it was finally installed :)

Though on another note, yes you are correct about the speed race. Just hope they keep the number of products down, as I can Imagine them doing 40 core versions @ 8ghz, 40 Core versions at 8.5ghz, then 60 Core versions of same values etc, could become a bit messy!
Jamie 18th January 2007, 13:15 Quote
I seems to me as though this multicore idea is taking a step in the wrong direction in terms of computer flexability.

If you specifically dedicate a core to hd video and suddenly nobody wants to do hd video does that core become redundant? If this is the case an out of date cpu will have lots of useless cores doing nothing.

Am I wrong?
Leitchy 18th January 2007, 13:16 Quote
Oh and does the software exist at the moment to dedicate a single/multiple core/s to a single application? With 80, it would be usefull assigning a certain ammount to say your favourite Image editing package so it has dedicated processing power.
Firehed 18th January 2007, 13:18 Quote
Quote:
Originally Posted by r4tch3t
You really don't like MS do you?

I think that this may be a step in the right direction, but it could just turn out to be the next GHz race.
You haven't tried running Vista on a single-core machine, have you? Mind you, it's not too bad on a dual-core rig, but even still feels a bit slow. OS X has spoiled me, I suppose. I haven't honestly paid a ton of attention to hardware since switching, but I don't think FrozenFire is lacking desperately by any standards (one gen behind on video cards I suppose, but that's about it), and Vista doesn't feel nearly as smooth as it should on that rig, even with a clean install. Although in hindsight, I think I've only tried RC copies. Umm... not that I've obtained one of the leaked final versions to compare it to, or anything.

But that aside, I'd love this. I really hate interacting with my fileserver, and that's solely because it's the only single-core machine I have left. Since getting a dual-core system, I've vowed to never use less again. I can only imagine how much better a quad would be not having used one, so this is just a mindgasm.

Jamie - I don't think that the cores are permanently dedicated to a specific task, I think that pic was just demonstrating how you could assign tasks and seemed to imply it would have a few reserved for such (not entirely unlike how Via CPUs have that hardware encryption that kicks the hell out of any madly clocked AMD or Intel chip by about an order of magnitude, with a quarter the clock speed). Of course, if you're talking 80 cores, I don't think having a few specialized would be a bad thing, especially since they tend to be massively faster at task X if it's designed solely for that purpose. You're talking no more than 5% wasted cores IF (!) that task was to go away forever, which seems pretty unlikely in the case of HD and graphics at least.
r4tch3t 18th January 2007, 13:19 Quote
Quote:
Originally Posted by Jamie
I seems to me as though this multicore idea is taking a step in the wrong direction in terms of computer flexability.

If you specifically dedicate a core to hd video and suddenly nobody wants to do hd video does that core become redundant? If this is the case an out of date cpu will have lots of useless cores doing nothing.

Am I wrong?
No your not, and I agree, with multipurpose CPUs they can adapt to suit any task, think of unified shaders.
Sure they can perform all of these tasks (possibly simulaniously, assuming sufficient bandwidth) But how much of a performance gain would you really get? They are dedicated but simpler. As far as I can tell they would boast no performance gain over traditional proccessors.
ch424 18th January 2007, 14:49 Quote
Jamie, I think the labels down the right hand side are examples of what each of the cores could be doing at once... HD video, crypto and graphics were just the buzzwords the art guy was given to play with ;) Each of the cores is the same, it's just that with 80 of them, you can assign loads of different tasks - it's making the point that you can assign many cores to a particular task with all those for GFX.

So yeah, what Firehed said...

ch424
Nexxo 18th January 2007, 15:28 Quote
80 cores? How big is that CPU going to be? The size of a motherboard? :p

It would work if they are basic CPU cores that can redistribute tasks on the fly, and configure the CPU array flexibly to suit the application. Ideally, to the software, the array would appear like a single CPU --just a really powerful and ideally suited one.
ch424 18th January 2007, 16:03 Quote
Nexxo, judging by the fact they have it running at 8GHz on current processes, at under 100W, it's not very big at all!
In terms of orders of ten, let's say current quad-core CPUs have 1 billion-ish transistors, and this has comparable thermal/electrical properties for 100ish cores, that means about 10 million-ish transistors per core... so Pentium II/III sort of level!

As for redistibuting tasks on the fly, that's presumably what these engineers are investigating.

There are other advantages of having 80 symmetric cores of course: you can put the manufacturing yeilds waay up by disabling the one or two cores that have defects, like they do with graphics chips (the X800GTO2 cards for example).

ch424
Gravemind123 18th January 2007, 16:10 Quote
So the next Celeron/Semprons will have cut cores instead of cache...that would be interesting...
I think this is a step in the right direction, but I'm not sure how many cores you will need, 80 just seems like overkill, but I'll have to see 5 years from now how things are.
Nexxo 18th January 2007, 16:44 Quote
Quote:
Originally Posted by ch424
Nexxo, judging by the fact they have it running at 8GHz on current processes, at under 100W, it's not very big at all!
In terms of orders of ten, let's say current quad-core CPUs have 1 billion-ish transistors, and this has comparable thermal/electrical properties for 100ish cores, that means about 10 million-ish transistors per core... so Pentium II/III sort of level!

As for redistibuting tasks on the fly, that's presumably what these engineers are investigating.

There are other advantages of having 80 symmetric cores of course: you can put the manufacturing yeilds waay up by disabling the one or two cores that have defects, like they do with graphics chips (the X800GTO2 cards for example).
That's the stuff. ;) Now if they can make the CPU array appear to software like a single really powerful CPU, we're in business. Don't program multithreaded apps; let the CPU array do it on the fly while the software is running.
WarMachine 18th January 2007, 17:29 Quote
Quote:
Originally Posted by Nexxo
That's the stuff. ;) Now if they can make the CPU array appear to software like a single really powerful CPU, we're in business. Don't program multithreaded apps; let the CPU array do it on the fly while the software is running.

I think that's the idea they should run with... I can see PC's looking very different in years to come. I envision being able to emulate operations and functions that are currently hardwired to the mobo with these processors, rendering the mobo to a simple socket and a few connections for hardware. Probably wouldn't be quite that simple, but I can see the potential to do so much with very little.
Nexxo 18th January 2007, 17:36 Quote
That is a very sexy idea... Perhaps they could even have a modular motherboard. A socket for the CPU, a few sockets for peripheral interfaces, NIC, perhaps a flash drive for the OS. All very compact. Nice.
ch424 18th January 2007, 18:05 Quote
What, like USB and PCI? :P
Anakha 18th January 2007, 19:09 Quote
Great. Except, of course, that these are "Simple" cores, which means RISC. Like the PowerPC chips, not like the x86 chips (Which are very CISC chips). It also means that the performance benefit with more cores would be a lot less. See, having a RISC processor would be like going back to a 286 (Or earlier) - no FP, no MMU, no SSE, no 3DNOW, no MMX - but clocked at modern (3GHz) speeds. More work would have to be done by in software to re-create the functions missing from the chip (Which software, especially multimedia software, demands these days).

As a (Totally inaccurate, but representitve) example, it'd be the difference between doing standard multiplication (9*9) and repeating the values and adding them up (9+9+9+9+9+9+9+9+9). In essence, they're the same action, but because the RISC processor is missing the functionality to do the * bit, it has to go the long way round, slowing everything, especially as instructions in software are an order of magnitude slower than those hard-coded into a chip (1 or 2 cycles per instruction for hard-coded, compared to 10-20 for software driven, what with all the memory reading and writing required).

And consider, Apple moved away from RISC/PowerPC, as the performance was nowhere near as good as it's CISC/x86 equivalent.
TomH 18th January 2007, 19:20 Quote
Quote:
Originally Posted by Nexxo
80 cores? How big is that CPU going to be? The size of a motherboard? :p

Well, TFA states that we'll be seeing these in 'five to eight' years. If we take that as being about 6 years (it's an easier number too), and look at how Moore's Law predicts CPU manufactuing process (halves in size every 18-24 months):

72 months, divided by 18 = 4 cycles of Moores Law.

So if we're at 65nm gate lengths now, we could be at 8.125nm in 6 years. :p

It's horribly inaccurate but the point is they'll be tiny compared with what we have now. Supposedly a decade ago, everything was built on 500nm processes!

There would be other huge factors involved in actually getting to that scale of manufacturing in reality, too.

Warning: You could get lost in this link for days :p
ch424 18th January 2007, 19:33 Quote
Anakha, it doesn't necessarily mean that: look at the 8800GTX. It has 128 "stream processors" all of which do insane amounts of FP work...
Tyinsar 18th January 2007, 19:35 Quote
Quote:
Originally Posted by Jamie
I seems to me as though this multicore idea is taking a step in the wrong direction in terms of computer flexability.

If you specifically dedicate a core to hd video and suddenly nobody wants to do hd video does that core become redundant? If this is the case an out of date cpu will have lots of useless cores doing nothing.

Am I wrong?
From the EETimes article linked in the story it looks like you are correct:
Quote:
...They're different kinds of cores, explains Vara...

...
While it may take five to eight years to come out with a working 80-core chip, Vara says IT managers might start watching for what he calls "different flavors" of quad-core chips. "Maybe you'll have interim chips where they have more complex cores along with simpler cores, too."
The "different flavors" worries me a little in terms of flexibility. I guess it all comes down to how versatile those different cores are. The odd thing is that, as I understand it, DX10 GPUs are actually getting more flexible chips than previous generations.
EQC 18th January 2007, 20:22 Quote
Quote:
Originally Posted by Tom Hill
Well, TFA states that we'll be seeing these in 'five to eight' years. If we take that as being about 6 years (it's an easier number too), and look at how Moore's Law predicts CPU manufactuing process (halves in size every 18-24 months):

72 months, divided by 18 = 4 cycles of Moores Law.

So if we're at 65nm gate lengths now, we could be at 8.125nm in 6 years.

I think the planned order is 65nm --> 45nm --> 32nm --> 22nm --> ??

And I'm not sure, but, I think the "halves in size" relates more to area than length...so 65/2 isn't equal to 45...but 65^2 = 2*45^2 (roughly). For the sizes in my memory (130nm, 90nm, 65nm, 45nm, 32nm, 22nm), the halving of "area" seems to hold more true than the halving of "length." If that holds true, then after 22nm we'll see something around 15 or 16nm.


Quote:
Originally Posted by Leitchy
Oh and does the software exist at the moment to dedicate a single/multiple core/s to a single application? With 80, it would be usefull assigning a certain ammount to say your favourite Image editing package so it has dedicated processing power.

Ummm, sortof. In my version of Windows XP-Pro (and on my old Hyperthreaded Intel Pentium 4, that looks like 2 processors to the OS), if I go into task manager (ctrl-alt-delete), and look at the programs running, I can right click and set the processor "affinity" for each one -- so I can make individual programs run exclusively on either of my virtual CPU's. I'd imagine you can do the same on true dual and quad-core processors.
monkeyville 18th January 2007, 22:41 Quote
This is very similar to the Cell Processor (the one in the PS3) that has 8 separate SPE's (cores). Considering that is in production now i can only see the cell producing the same kind of cores in much short a time frame. I was under the impression that task specific processors/cores required much greater time to code though which could be a hindrance.
TomH 21st January 2007, 03:30 Quote
Quote:
Originally Posted by monkeyville
I was under the impression that task specific processors/cores required much greater time to code though which could be a hindrance.
Bit ran an article about Valve's implementation of multicore support in Source, which goes into quite a bit of detail about dedicating threads for different 'areas' of the engine (physics, audio etc.) and how they went about coding it.

Found it pretty interesting myself :)
Quote:
Originally Posted by EQC
I think the planned order is 65nm --> 45nm --> 32nm --> 22nm --> ??

And I'm not sure, but, I think the "halves in size" relates more to area than length...so 65/2 isn't equal to 45...but 65^2 = 2*45^2 (roughly). For the sizes in my memory (130nm, 90nm, 65nm, 45nm, 32nm, 22nm), the halving of "area" seems to hold more true than the halving of "length." If that holds true, then after 22nm we'll see something around 15 or 16nm.
Ok, well technically Moore's Law states that every 24 months, the amount of transistors on a die will double. For this to happen, it's my understanding that the transistors ideally have to become half of the size they were 2 years hence. So as you've pointed out, die size does come into it too.

If you plotted date/process size, you'd see a pretty strong correlation along x=y, but it's obvious that the Law isn't much of a Law, as instead a general rule of thumb. There's something like this on the Wiki article, which better explains it.

It's been sagging under expectation, but climbing back again. :)
sinizterguy 21st January 2007, 12:52 Quote
Without a complete OS and software overhaul, these processors will not make it into the general consumer market.

While Windows Vista may be multi-threaded it damn sure is not massively multi-threaded enough to make of more than about 4 cores or so.

Then all programs will have to almost completely re-written to actually make use of this kind of processor - or it will just end up running slower than before.

So, I think these might make it into server markets, but not consumer markets.
EQC 23rd January 2007, 08:40 Quote
Quote:
Originally Posted by Tom Hill

Ok, well technically Moore's Law states that every 24 months, the amount of transistors on a die will double. For this to happen, it's my understanding that the transistors ideally have to become half of the size they were 2 years hence. So as you've pointed out, die size does come into it too.

If you plotted date/process size, you'd see a pretty strong correlation along x=y, but it's obvious that the Law isn't much of a Law, as instead a general rule of thumb. There's something like this on the Wiki article, which better explains it.

It's been sagging under expectation, but climbing back again. :)

Thanks for the further info...I got more interested, since the image you linked didn't go far enough to include Core2Duo...so I actually looked up some more numbers. Interestingly, I found that the Pentium D 900 series had 376 Million transistors, while the Core 2 Duo dropped back down to 291 Million. Of course, the Core 2 Quad jumps back forward to 582 Million transistors.

I found the info in the tables on this and this page.
Tyinsar 23rd January 2007, 21:45 Quote
Thanks for the links EQC, You know, I had seen that before but it keeps amazing me that AMD is using 154M transistors on their X2s and ahead of the Pentium D which has more that twice as many and while the C2D is faster it still uses almost double - No I'm not an AMD fanboy, my last CPU purchase was an e6300, but I am still impressed by AMD for this.
matt.slysnake 23rd January 2007, 23:14 Quote
80 cores.. wow, will there ever be a limit?
brumster 24th January 2007, 03:00 Quote
I will never need more than 640k cores.
Tyinsar 24th January 2007, 03:15 Quote
Quote:
Originally Posted by brumster
I will never need more than 640k cores.
:)
Log in

You are not logged in, please login with your forum account below. If you don't already have an account please register to start contributing.



Discuss in the forums