GlobalFoundries claims high-k advancement

[Idontcare]

GlobalFoundries claims high-k advancement

At the 2009 Symposium on VLSI Technology in Kyoto, Japan, GlobalFoundries Inc. claims that it has found a technique that enables a high-k/metal-gate transistor to scale to the 22-nm node and beyond.

The foundry vendor reported the first demonstration of a technique that allows the equivalent oxide thickness (EOT) in a high-k/metal-gate transistor to scale. The results were demonstrated through the fabrication of an n-MOSFET device with EOT of 0.55-nm and a p-MOSFET with EOT of 0.7-nm.

http://www.eetimes.com/news/se...l;?articleID=217801312

Pretty cool to see 0.5nm EOT demonstrated.

My understanding is that Intel will, at some point in the conference, be talking about gate stacks targeted at the 16nm node which will entirely eliminate the silicon oxide interface currently found in all high-k gate implementations.

 

[Cogman]

Ahh, and they said it couldn't be done

 

[thilanliyan]

Something by Toshiba:
http://www.digitimes.com/news/a20090615PR202.html

 

[heyheybooboo]

Here's hoping GF is around at 22nm

 

[Idontcare]

Originally posted by: thilan29
Something by Toshiba:
http://www.digitimes.com/news/a20090615PR202.html

Interesting, they are using Germanium for the channel there.

 

[Nemesis 1]

Dam 2 years after Intel does Hk /Metal gates and 2 years After IBM/AMD SAID we have that to . You just wait and see. Still waiting. Now we get announcement about AMDs improved version at 22nm . I still want to see the orginal unimproved AMD highK / Metal gates. OR is AMD improving for intel . and spending research dollars. AMD is talking way to much about 22nm . If they can get there first shut up fools.

 

[Hacp]

Originally posted by: Nemesis 1
Dam 2 years after Intel does Hk /Metal gates and 2 years After IBM/AMD SAID we have that to . You just wait and see. Still waiting. Now we get announcement about AMDs improved version at 22nm . I still want to see the orginal unimproved AMD highK / Metal gates. OR is AMD improving for intel . and spending research dollars. AMD is talking way to much about 22nm . If they can get there first shut up fools.

isn't intel going to immersion on 32nm?

 

[Nemesis 1]

Thats the way I have always understood it , But ya know I never ever did check, nor do I recall what intel did say . But I think ya. It really bothers me AMD is talking about 22nm . So much.

I half exspected AMD to beat Intel to 32nm. I have a nice 2 core chip here. Really I don't like what intel is doing right now. Don't you think intel is being terriably quit about chips that are basicly ready . M/B are ready. Intel is impressive but they cornered themselves.

 

[CTho9305]

What are the equivalent oxide thicknesses for various modern processes?

 

[soccerballtux]

HfO2 on the gate is 80% the thickness of SiO2.* for Intel
IIRC.
I recall thinking "that's odd they'd go to HfO2 if the improvement is only 20%"

 

[Idontcare]

Originally posted by: CTho9305
What are the equivalent oxide thicknesses for various modern processes?

Intel's is 1nm EOT for 45nm, for AMD I believe the stated number is 1.2nm but I can't locate a handy public reference. Will keep looking.

 

[deputc26]

Originally posted by: Hacp

Originally posted by: Nemesis 1
Dam 2 years after Intel does Hk /Metal gates and 2 years After IBM/AMD SAID we have that to . You just wait and see. Still waiting. Now we get announcement about AMDs improved version at 22nm . I still want to see the orginal unimproved AMD highK / Metal gates. OR is AMD improving for intel . and spending research dollars. AMD is talking way to much about 22nm . If they can get there first shut up fools.

isn't intel going to immersion on 32nm?

That is correct

 

[Idontcare]

Originally posted by: Nemesis 1
Dam 2 years after Intel does Hk /Metal gates and 2 years After IBM/AMD SAID we have that to . You just wait and see. Still waiting. Now we get announcement about AMDs improved version at 22nm . I still want to see the orginal unimproved AMD highK / Metal gates. OR is AMD improving for intel . and spending research dollars. AMD is talking way to much about 22nm . If they can get there first shut up fools.

Having it is one thing, a matter of nothing more than technical capability and engineering.

It being the most cost-effective pathway to creating a xtor and IC with the desired performance attributes is entirely a whole other thing.

In 2004 TI had both HK for the gates and ULK for the BEOL fully developed and yielding well enough to make sellable product on 65nm tech. But the process technology required to do so cost an absolute bloody fortune, ergo we didn't put it into production but rather shelved it for 45nm and 32nm usage.

I have no doubt Intel had HK/MG tech ready and usable for 65nm as well, but the costs likely did not justify the performance gains...they could get that performance more cheaply by using traditional scaling and materials.

I view the IBM fab club situation similarly...sure they had the technology, I absolutely believe them (even more so because I know a number of their R&D folks) but I also believe it wasn't the lowest cost pathway for them to hit their electrical specs (Idsat, Idrive, Ioff, Iddq, etc) so they took the more traditional approach in the production node.

I know you have an emotional/psychology issue with IBM and any business that associates with them, so your comments regarding them needing to shutup and being fools are not unexpected, but if you tone down the bias dial just a bit and ask yourself from a shareholder standpoint what would you like to see IBM and the others do here I think you'd be just fine with them publicizing R&D technical achievements in an attempt to maximize the shareholder value of their patent portfolio.