Hi Miroslav,
I encounter some issues compiling your exciting code for Windows x86. I've compiled with Code Blocks under GCC 4.8.1 and have a SEG FAULT when I use SSE, ie -F4 and -J7. Seems to be at pthread_create function. I'd linked memalign to _aligned_malloc, buffers seems to be initialized.
I' don't know how to make it works on windows. Have you an idea ?
Thanks... Guilain