AMD Says That CPU Core Race Can't Last Forever
128 cores? Sadly, no.
It wasn't too many CPU generations ago that the main focus of performance was clock speed. The public perception was that the more megahertz (or gigahertz), the better.
Now, the current competitive number appears to be in the number of cores inside a processor. Donald Newell, AMD's chief technology officer for servers, believes that this sort of race to an even greater number of cores cannot continue.
Interestingly enough, Newell knows all about the numbers game as he previously spent 16 years at Intel, during which time there was the very clock-happy Pentium 4 generation.
"We thought we were going to build a 10GHz chip. It was only when we discovered that they would get so hot it would melt through the Earth, that we decided not to do that," Newell said, jokingly, in an interview with IDG.
Now it's about who has more cores, but Newell doesn't see that continuing indefinitely.
"There will come an end to the core-count wars. I won't put an exact date on it, but I don't myself expect to see 128 cores on a full-sized server die by the end of this decade," said Newell. "It is not unrealistic from a technology road map, but from a deployment road map, the power constraints that people expect [servers] to live in."
While we haven't seen the end of core count growth, the next big competitive ground could be integrating specialized functions into the processor.
"There is nothing to prevent us to put specific features on die that enable more efficient processing," Newell said. "So you should expect to see heterogenous architectures to emerge where we identify functions that are broadly useful but don't necessarily map into an instruction that you'd want to add directly into the x86 architecture."
Both AMD and Intel are integrating graphics components into their processors, but AMD's Fusion solution promises to be the more capable offering with greater power available for GPGPU functions.
Source: IDG.
- AMD to Make a Tablet Chip in the Future
- AMD/ATI Ships 25 Millionth DirectX 11 GPU
- Microsoft Uses Old Spice Man for Win Phone 7 Ads
- Video Game Sales Fall for 6th Consecutive Month
- Iomega Reveals USB 3.0 External SSD
- Congrats To Our Intel K-Series Contest Winners
- AMD Announces Net Loss in the Third Quarter
- IBM Making Nvidia Fermi-filled BladeCenter Server
- PC Sales Are Up, Macs Getting Even More Popular
- Tim Willits: RAGE 2 Side Games Possible
- Is This the New, Even Thinner MacBook Air?
- Microsoft Gives More Free Software to NGOs
- Apple Making Huge Money; iPad Outsells Macs
- Steve Jobs Hates on RIM, Android, Smaller Tablets
- WD's 3TB Internal HDD Comes with PCI-e Card
- Blizzard Sues StarCraft II Hackers
- Mimedia Online Storage Sends HDD For Backup
- Microsoft's Chief Software Architect Quits






Thats because AMD can never beat intel!!lol!!
Wel las any ATI fan would tell you, "beating is not by performance but by power saving" ...
Personally speed comes before power, as you can do things faster and then ... shock horror ...turn it off! lol
Until the equipment is ridiculously fast, more cores + more power is what i will be buying, we can worry about efficiency later eh ^^
Thats because AMD can never beat intel!!lol!!
Yeah, I guess the entire K8 thingy never happened...
Seriously, Intel took a short-term lead, but in a world that changes as fast as computing, AMD could crush Intel in less than a year and vice versa.
I dont think thats likely Silmarunya, Intel have more resource and could last years of performing badly ... fortunately for them though, they do not need to "survuve" they have millions of people that trust their kit and better performance to boot. (not better consumption i know).


AMD are hardly likely to "kill" anyone ...even if they do things perfect ..but i do welcome the competition,
Be aware we are on planet earth, and this is not CPU deathrace 2000 (fanboiwars)
Why people keep thinking this is some unrealistic war to the death will allways bemuse me (perhaps watch to many films?)... If anyones gonna have trouble look at the people struggeling to make profit
I think Silmarunya was just referring to the the fact that AMD could overtake Intel in performance stakes in a short period of time, and vice-versa if AMD were currently on top - which is entirely accurate and has happened in the past both ways. Not that AMD could bring Intel to bankruptcy.
You're either too young to know what you're talking about or have a very short memory if you think any tech company has some infallible position.
They havn't even approached the performance of intel yet atall!.



I am 33 years old and have upgraded my PC's since my first 33mhz 486 DX2 (66Mhx in turbo mode).
AMD are struggeling to make a profit and the only thing they have going for them in the CPU market at the moment is:
1) power effieciency (good "bang for buck")
2) fanboi'ism (alot of people who stick up for them even though the hardware is not as good as the competition, aka: fans)
3) purchases on a budget (cost is awesome)
If you think that's enough to take over the world then go for it ...you'll be trying some time
I got what Sil' was reffering to, i just dont think it's that relevant (referring to complete dominance since day1 as "short term" ??) ...thanks!
Now can we state different opinions on here withought being presumed ill informed or too young please?
I dont think thats likely Silmarunya, Intel have more resource and could last years of performing badly ... fortunately for them though, they do not need to "survuve" they have millions of people that trust their kit and better performance to boot. (not better consumption i know).AMD are hardly likely to "kill" anyone ...even if they do things perfect ..but i do welcome the competition,Be aware we are on planet earth, and this is not CPU deathrace 2000 (fanboiwars) Why people keep thinking this is some unrealistic war to the death will allways bemuse me (perhaps watch to many films?)... If anyones gonna have trouble look at the people struggeling to make profit
1) In what universe does Intel lose on power consumption? It's easy to criticise Intel's architectures, but power efficiency is not amongst its weaknesses. Partly due to better architecture and partly due to die shrinkages, Intel CPU's consume far less power than comparable AMD competitors (95W for the i5-750 vs 125W for AMD's 965 for example, and that i5 has an on board northbridge too).
2) Of course AMD won't kill a company with such a long history, brand name recognition and cash reserve as Intel, where on earth did I suggest that?
3) All I was saying is that in a field where architectures shift every 2-3 years, it's very hard to say a company is 'winning' or 'losing'. Take a look at the GPU market for example. ATI's 3000 series had to play second fiddle right out of the gate, the 4000 was decent enough to force Nvidia into restructuring their prices and the 5000 crushes Nvidia outright. Same thing happened when Intel replaced the dead horse called Netburst with their newer models. Same thing happened when AMD released K8. Same thing happened when Windows came up with Windows 7 vs Vista or XP vs ME.
Well then assuming you're not able to remember back to the K8/P4 days, where AMD was running circles round Intel, then perhaps you can remember a few years back where nVidia was basically rebadging the 8800GT over and over again since AMD/ATI couldn't compete with any of their new cards, and how that flipped round the situation we have now where AMD is streets ahead of nVidia again.
It's entirely possible in the space of a year for AMD to be ahead of Intel again, and my last three upgrades have been Intel based, so this isn't fanboyism speaking, it's experience and realism.
Hi sil'
1) thats what all the AMD fanbois keep on assuring me everytime i compliment my own harware, more of a general aim (not at you). i dont see a saving myself either.
If you have something that takes half the power but takes twice as long, it's not a power saving ...this i know.
2) this point is part of point 3,
3) the CPU industry has been going far longer than 2-3 years ...look at the long term and you will see why i consider 2-3 years as "insignificant" for intel.
Hope that helps clear it up, not looking for a fight here, just think the point was well ... pointless
Well then assuming you're not able to remember back to the K8/P4 days, where AMD was running circles round Intel, then perhaps you can remember a few years back where nVidia was basically rebadging the 8800GT over and over again since AMD/ATI couldn't compete with any of their new cards, and how that flipped round the situation we have now where AMD is streets ahead of nVidia again.
It's entirely possible in the space of a year for AMD to be ahead of Intel again, and my last three upgrades have been Intel based, so this isn't fanboyism speaking, it's experience and realism.
I remember them days tyvm ... they were the days AMD were running on VIA motherboard chips and failures were far to common! (hence a reason to avoid them).
And i disagree with your last paragraph .. that's not happening ... i guess you have to wait to prove me wrong, right?
Hi sil'1) thats what all the AMD fanbois keep on assuring me everytime i compliment my own harware, more of a general aim (not at you). i dont see a saving myself either.If you have something that takes half the power but takes twice as long, it's not a power saving ...this i know.2) this point is part of point 3,3) the CPU industry has been going far longer than 2-3 years ...look at the long term and you will see why i consider 2-3 years as "insignificant" for intel.Hope that helps clear it up, not looking for a fight here, just think the point was well ... pointless
AMD has had a lead over Intel for a significant period of time. The entire lifespan of the K8 architecture was an era in which AMD crushed Intel in everything - performance, price and power consumption. The tables are turned right now and they will be turned over and over again for years to come. Same thing can be said of just about every sector.
In the GPU market, Nvidia and ATI keep taking the lead only to give it back at the next architecture. Windows, Ubuntu and several other OS makers keep releasing mediocre or downright poor products only to be followed by excellent ones and so on.
Hardware and software design is an eternal game of leapfrog, denying that is the same as denying you and I are human beings.
I didnt deny leapfrogging Sil ....Also please look at the word "Crush" in the dictionary (the word you used the first time). Maybe you used the wrong one?

...I did state no-one will kill anyone and competition is "good and welcomed" ...sheesh you people are jumpy ...got problems?
I didnt deny leapfrogging Sil ....Also please look at the word "Crush" in the dictionary (the word you used the first time). Maybe you used the wrong one? ...I did state no-one will kill anyone and competition is "good and welcomed" ...sheesh you people are jumpy ...got problems?
Maybe your reading comprehension is lacking, but in this context crushing meant releasing a more powerful and power efficient product, not driving the competitor bankrupt, as malphas correctly assumed.
They need to work out how to increase the Ghz, so the speed race can start rather than core race.
Graphene chips please!
Well that's what you get for leaving things down to assumption and your "version" of the context i guess Sil' ...misunderstanding, you guys could still take less offence to it yknow tho?
And well said Lewis ^^
A pathetic excuse! I want a 10GHz chip, even if I have to build an extension onto my flat to accommodate the heatsink! :-)
A pathetic excuse! I want a 10GHz chip, even if I have to build an extension onto my flat to accommodate the heatsink! :-)
Don't worry, once they get to organic transistors, you might get what you want. Just wait a few years and start auctioning some organs.
Besides, why is everyone so obsessed with increasing clock speed? If software is properly optimised, multiple cores are often better than higher clock speed. And actually, architecture > clock speed.
Personally, I'd rather have GPU-assisted parallel computing pretty soon. I'm not an Nvidia fanboy, but that's something they are 100% spot on.
Harnessing the gpu whilst its idle is a great idea, so is the efficiencys of multithreading, when all my cuda cores (and for other their stream processors) and also my quad hyperthreaded 32 core x4 die's with quad channel memory (on 4 seperate buses) meresistors for my ssd and lightpeak fpr transfer hit the desk i will be rather pleased...
...might have to wait sometime ...the ride is nice tho eh?
Harnessing the gpu whilst its idle is a great idea, so is the efficiencys of multithreading, when all my cuda cores (and for other their stream processors) and also my quad hyperthreaded 32 core x4 die's with quad channel memory (on 4 seperate buses) meresistors for my ssd and lightpeak fpr transfer hit the desk i will be rather pleased......might have to wait sometime ...the ride is nice tho eh?
Why not aim a bit higher? I mean, why stop there when you could have a pocket sized, 25 petaqubit quantum computer?
Why not aim a bit higher? I mean, why stop there when you could have a pocket sized, 25 petaqubit quantum computer?
who said i was stopping? ...i was talking about the next time i'll upgrade
see what happens when you rely on assumption