Future processors under 100nm
This article talks about how as processors are made smaller and smaller under 100nm, they have more electrical leakage between gate and such that they will dissipate more heat than usual.
Estimates are over 100 watts! And the measures proposed to combat the heat issue are complex and wasteful at best. Hopefully IBM or some other comapny will come up with a method to control this leakage but so far it's still an issue. What's the purpose of investing in a smaller process if it will use MORE electricity and create MORE heat than modern 130nm processors? 90nm may be the end of the road for the time being until these problems are resolved or the research is turned over to some more exotic chip development like optic circuitry, organic-bio circuits or something else.
Estimates are over 100 watts! And the measures proposed to combat the heat issue are complex and wasteful at best. Hopefully IBM or some other comapny will come up with a method to control this leakage but so far it's still an issue. What's the purpose of investing in a smaller process if it will use MORE electricity and create MORE heat than modern 130nm processors? 90nm may be the end of the road for the time being until these problems are resolved or the research is turned over to some more exotic chip development like optic circuitry, organic-bio circuits or something else.
Comments
IBM expects the Fishkill plant to make it to 65nm and possibly beyond. There's still plenty of time for some clever team of engineers to route around the problem. They've done it before.
I don't think that the 90 nm process will be an issue. Intel is ready to move on this process this year, and IBM the next year. 65 nm process should be an harder problem.