Hello Computing Performance,
I am reading the MK book "Intel Xeon Phi Coprocessor High Performance Programming", and I just realized Intel has two whitepapers on Xeon Phi tuning already up on their site:
The MIC architecture is interesting mostly because optimizing for a 10-core Xeon or a 100-core coprocessor works along the same lines, and tuning performed to increase parallelism for the coprocessor will benefit execution on a standard Xeon CPU as well -- given how time-expensive code optimization is, this is highly interesting.
Best -F
_________________________________________ -- "'Problem' is a bleak word for challenge" - Richard Fish (Federico L. Lucifredi) - flucifredi at acm.org - GnuPG 0x4A73884C