Re: Intel Compiler Reimbursement Fund
- From: "FX" <coudert@xxxxxxxxxxxxx>
- Date: Sun, 6 Feb 2011 09:37:34 +0000 (UTC)
It was, and still is, the Math Kernel Library that has the speed bump.
I don't know whether the MKL had it, but the runtime library of the
compiler itself happily performed this naughty check, as explained for
example in http://www.swallowtail.org/naughty-intel.shtml (the vectorized
math library was part of the compiler runtime, not MKL).