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).
gpu dynamic opencl zero-copy pool load-balancer parallelism gpgpu gpu-acceleration opencl-kernels gpu-computing pipelining batch-processing multi-device multi multi-gpu iterative
-
Updated
Jul 7, 2022 - C#