3 ms·
What’s Next for Moore’s Law? For Intel, III+V = 10nm QWFETs
- zokier 12y agoI don't know much about semiconductor manufacturing, but moving to QWFETs and completely new materials seems like huge change for something that supposedly is going to happen in the next 6-12 months. If I'm not mistaken this would be one of the biggest changes in semiconductors in the past 10-20 years?
- mscrivo 12y agoAs noted in the article, Intel has been working on this for just under a decade, but still, it does seem highly optimistic.
- minthd 12y agoFinfet was also really hard , and took something like a decade(or more) from research to production with decent yield. And EUV based lithography(printing finer circuits using xrays), which intel recently bought 10-20 machines of, also took a very long time. So without a doubt, making chips is becoming much harder.
- ipsum2 12y agoDo you know who they bought the EUV machines from? I didn't know EUVs are producing high enough yields yet to be used for commercial applications.
- minthd 12y agoThey bought it from ASML. Yield wasn't the problem, throughput was - and many experts lost hope on EUV - so this purchase is quite good news.
- Detrus 12y agoNo every shrink involved some material changes.
- AnonNo15 12y agoHere is the source paper he might be referring to. https://www.intel.ca/content/dam/doc/white-paper/non-planar-multi-gate-ingaas-qwfet-5nm-paper.pdf https://www.intel.ca/content/dam/doc/white-paper/non-planar-...
- ableal 12y agoThe RWT site is a gold mine of information and analysis of CPU history, for instance this 2006 piece: http://www.realworldtech.com/vax-cpu-economics/ http://www.realworldtech.com/vax-cpu-economics/ Linus Torvalds used to drop in to comment, a few years ago.
- berkut 12y agoHe still does.