Mir GLAS vs Intel MKL: which is faster?

[Andrei Alexandrescu via Digitalmars-d]

On 09/24/2016 01:46 PM, Ilya Yaroshenko wrote:
> Eigen was added (but only data, still need to write text). Relative
> charts was added.

Looks awesome. Couple more nits after one more pass:

"numerical and scientific projects" -> "numeric and scientific projects"

"OpenBLAS Haswell computation kernels" -> "The OpenBLAS Haswell 
computation kernels"

"To add a new architecture or target an engineer" -> "To add a new 
architecture or target, an engineer"

"configurations are available for X87, SSE2, AVX, and AVX2 instruction 
sets" -> "configurations are available for the X87, SSE2, AVX, and AVX2 
instruction sets"

In the machine, you may want to specify the amount of L2 cache (I think 
it's 6 MB)

Instead of "Recent" MKL, a version number would be more precise

Relative performance plots should specify "percent", i.e. "Performance 
relative to Mir" -> "Performance relative to Mir [%]"

"General Matrix-matrix Multiplication" -> "General Matrix-Matrix 
Multiplication"


Andrei