Matrix multiplication. More work per thread v.3 benchmark

WebGL2-compute based NxN matrix multiplication C = A x B Shader v.3 benchmark. Fixed random A, B 4096x4096 buffers are used for calculations, therefore You can set new N value (up to N=4096, multiple of TS). TS - tile sizes (local_size_x,y). it - number of iterations. FLOPS = 2 N3it / time. Random CGPU and CGPU - CCPU are shown.

N= TS= it=

GeForce GT 710     ~54 GFLOPS (N=4096, TS=32).
Intel Atom Z3735F   8.5 GFLOPS (N=1024, TS=32).

SGEMM in WebGL2-compute     updated 26 Mar 2019