tugrul512bit / Cekirdekler

Multi-device OpenCL kernel load balancer and pipeliner API for C#. Uses shared-distributed memory model to keep GPUs updated fast while using same kernel on all devices(for simplicity).
GNU General Public License v3.0
93 stars 10 forks source link

Complete device to device pipeline stage initialization kernel execution #31

Closed tugrul512bit closed 7 years ago

tugrul512bit commented 7 years ago

so it doesn't start with garbage, especially for hidden buffers per stage (hiddens also needs testing)