3 ms·
If it something like OpenMP (meta language on top of general purpose one) then I can see it, but if it's something that forces one to write always AMP code in t
by malkia 15y ago
If it something like OpenMP (meta language on top of general purpose one) then I can see it, but if it's something that forces one to write always AMP code in the same language, then it would get messy very fast. I much preffer #pragma opm parallel for, rather than parallel_for<> and some syntax that would try to get into the original language.
Why Microsoft can't reuse OpeNCL?
- sc68cal 15y agoThe same reason they introduced DirectX to compete with OpenGL. Lock-in.
- chaffy 15y agoMaybe, but to be fair, that's only part of it. OpenCL is C99-based, and MS wants a C++ solution. OpenCL introduces more problems than it solves, primarily because the standards body mandated C99 and a low-level, driver-style interface. Have you ever seen an OpenCL Hello World program [1]? It's miserable. [1] http://developer.apple.com/library/mac/#documentation/Performance/Conceptual/OpenCL_MacProgGuide/Example:Hello,World/Example:Hello,World.html http://developer.apple.com/library/mac/#documentation/Perfor...
- malkia 15y agoCan you show us a DirectCompute one? Btw, introducing C++ introduces problems too: RTTI, exceptions, and the fact that templates are to be used (would tend to create bigger code).
- chaffy 15y agoDirectCompute suffers much of the same problems as OpenCL. Folks who want to program GPUs productively use Thrust [1]. [1] http://code.google.com/p/thrust/ http://code.google.com/p/thrust/