Udhay Shankar N <[EMAIL PROTECTED]> writes:
> http://www.mcgill.ca/newsroom/news/item/?item_id=102307

Ignoring the article (which seems like the typical breathless press
release by a researcher looking for money that will lead to nothing)...

Lets just address the "Moore's law" question for a moment.

Shrinking transistors clearly has a limit -- we aren't going to get
switching elements much smaller than a small molecule. Call it "one
atom" as a reasonable floor. That isn't so many halvings of size away
from where we are. However, right now, circuits are all 2D, and moving
to 3D will give us plenty of room to do fun new things for some time.

More interesting right now is the fact that we've hit something of a
wall on processor clock speed -- for the near future, speed
improvements are going to come from parallel architecture. It isn't
clear we know how to exploit parallel architectures terribly well.

Perry

Reply via email to