Credit to Multicoreinfo.com for the pointer to this article in the EE Times about the fruits of some of Dr. Reed’s travels
The U.S. needs to double its spending on basic research in physical sciences, according to Daniel A. Reed, a member of PCAST, the President’s Council of Advisers on Science and Technology. Reed testified before Congress on the issue recently and spoke with EE Times at the official opening of the Berkeley Parallel Computing Lab here.
Predictions?
In the short term, Microsoft will roll out a functional language called F# as part of its Visual Studio tools, Reed said. The effort is the nearest to commercialization of many parallel software projects at Microsoft, he added.
In terms of hardware, Reed predicted hybrid processors will emerge during the transition. They will include a mix of a few large, out-of-order execution cores to handle existing serial code as well many small in-order cores to handle parallel tasks.
I was able to watch the beginning of the video from the opening of the Berkeley UPCRC in a meeting break. He’s a lot better than I am at extemporaneous video.
