Intel announcement’s last week about its upcoming 45-nm CPU with “high-k” was hailed by the mainstream press as a “breakthrough.” IBM’s disclosure about its high-k material developed with
AMD was given less attention, but it too was about upcoming 45-nm CPUs that will become available in 2008.
The announcements were a big deal, but is “breakthrough” the right word to describe them? Intel says high-k dielectrics in conjunction with 45-nm transistors will offer a 30% reduction in switching power—thermal issues have been a huge problem with Intel’s devices during the past few years. Intel says the new design will boost switching speed by 20% and, especially, offer a 10-fold decrease in transistor gate oxide leakage, which indeed represents a major development.
I don’t think, though, that the news will affect our computing experiences involving most of what the average person does, namely office apps, Internet access and email. An AMD spokesman I contacted agreed. While the development will enable CPU makers to proceed along their same historical path of keeping up with Moore’s law, it remains to be seen what, if any resulting applications will really change our computing experiences that much. I mean, my processor will run so much faster next year; but will I really notice a difference?
