AMD and IBM announce 24 percent speed breakthrough

[OzzieGT]

http://www.pcpro.co.uk/news/66...peed-breakthrough.html

Since it's short, here is the whole article:

"A new faster version 'strained silicon' has been developed by IBM and AMD. The companies are claiming that the resulting transistors are up to 24 per cent faster than existing processes.
The companies say that the new strained silicon process, named `Dual Stress Liner,` improves the performance of both n-channel and p-channel types of transistors by stretching the silicon atoms in one transistor and compressing them in the other.

As the speed increase occurs without any increase in power, the partners foresee a performance boost without incurring the kind of problems with heat dissipation which dogged Intel's attempts to increase the clock speed of its Pentium designs to 4GHz.

As an added bonus, dual stress liner can be introduced without massive investments in new production techniques and can be integrated into commercial production lines rapidly.
AMD says that it aims to introduce the new strained silicon technology into all of its 90nm processor platforms as soon as the first half of 2005. Among the chips to include dual stress liner technology will be multi-core AMD64 processors.

Further details will be announced at this week's IEEE International Electron Devices Meeting in San Francisco."

 

[Brian23]

Anand said it was 12%

 

[DAPUNISHER]

Originally posted by: Brian23
Anand said it was 12%

Where? All I see is that blurb from Adul, and they corrected that in the feedback by pointing out that 24% is what is claimed by the press release.

 

[StealthElephant]

CNN.com article said 12% as well.

 

[klah]

http://www.amd.com/us-en/Corpo..._104_543~91999,00.html

The breakthrough process results in up to a 24 percent transistor speed increase, at the same power levels, compared to similar transistors produced without the technology.

 

[Mik3y]

woohoo for amd!!!

 

[Zebo]

i think the confusion stems from FX-55's improvement plus this lastest delevopment add up too 24% which will be seen on the new 90nm which can already do 2600Mhz... meaning 3250Mhz should'nt be too hard to expect...

so prolly 12 and 12?

 

[BW86]

cool

 

[Joerg]

So in other words we might see a dual core AMD 5000+ in reviewers hands by june or July? That would be a pretty amazing breaktrhough if so. Good job AMD.

 

[Mik3y]

Originally posted by: Joerg
So in other words we might see a dual core AMD 5000+ in reviewers hands by june or July? That would be a pretty amazing breaktrhough if so. Good job AMD.

nope. that's way too early. give or take a year from now for dual-core to come out to average consumers. they will primarily be implemented for servers and workstations at first. dual-core will primarily be based off socket 900, though i've heard that some will work with socket 939 with a bios flash.

 

[bob661]

off socket 900

Actually, AMD is debating on what socket to use for the desktop chips. They're thinking of going to a 941 or 942 but they don't want to piss of their partners so it might just be 939. The Opterons will stay on 940 with the dual cores as they've already promised that only a BIOS upgrade will be needed for present boards to accept dual cores. Personally, I don't see why a new socket would be needed for 939 chips when 940 chips will have dual core support.

 

[Mik3y]

Originally posted by: bob661

off socket 900

Actually, AMD is debating on what socket to use for the desktop chips. They're thinking of going to a 941 or 942 but they don't want to piss of their partners so it might just be 939. The Opterons will stay on 940 with the dual cores as they've already promised that only a BIOS upgrade will be needed for present boards to accept dual cores. Personally, I don't see why a new socket would be needed for 939 chips when 940 chips will have dual core support.

same. everything i'm hearing are pretty much rumors, so i'm not sure if it's 100% true. your reason sounds logical also, but i think amd wants to change from 940 because of the registered memory.

 

[jiffylube1024]

Originally posted by: Joerg
So in other words we might see a dual core AMD 5000+ in reviewers hands by june or July? That would be a pretty amazing breaktrhough if so. Good job AMD.

I think both companies have said that initial dual-core chips will be a few speed-grades down from top-of-the-line. Still, even dual 3800+ or 4000+ chips would be extremely impressive next summer.

 

[stelleg151]

umm, what does a change in silicon have to do with dual core, the impressive thing is 3.25ghz single core.

 

[JBT]

Damn nice I think I might need to be investing some money in AMD...

 

[Bateluer]

Originally posted by: stelleg151
umm, what does a change in silicon have to do with dual core, the impressive thing is 3.25ghz single core.

Because a dual core 2.6Ghz chip would likely outperform a single 3.25Ghz single core, especially in any multi-tasking intensive tasks. Now, AMD needs to implement a version of Hyperthreading into their cores. Dual core Athlon 64 with hyperthreading, now that'd be sweet.

 

[clarkey01]

that might be hard Bateluer, for AMD to have a type of hyper-threading in thier cores, due to thier short pipeline. (12 stages for the K8, 31 for the P4).

 

[StrangerGuy]

I want my 3GHz A64 now!

 

[bob661]

logical also, but i think amd wants to change from 940 because of the registered memory.

940 is pretty much going to stay that way because that's their server platform. And the enterprise environment does not like change.

 

[jiffylube1024]

Originally posted by: Bateluer

Originally posted by: stelleg151
umm, what does a change in silicon have to do with dual core, the impressive thing is 3.25ghz single core.

Because a dual core 2.6Ghz chip would likely outperform a single 3.25Ghz single core, especially in any multi-tasking intensive tasks. Now, AMD needs to implement a version of Hyperthreading into their cores. Dual core Athlon 64 with hyperthreading, now that'd be sweet.

AMD doesn't need HT when they go dual core, especially not with such a short pipeline.

 

[FlameDeer]

Great breakthrough! A new hope for AMD!

 

[IntelUser2000]

24% over no strained silicon. 12% over strained silicon previous gen. AMD also stated dual-cores start intially at lower clock speeds at least for Opterons.