Hi,
I tested my code in 2 differents GPUs :
- GeForce 9300M GS (OpenCL 1.0) - GTX 580 (OpenCL 1.1)
When I execute the code show bellow, in 9300M card its more fast than gtx 580:
GPUDevice.oclCQ.Write<byte>((CL.ComputeBuffer<byte>)Memory.Mem, true, 0, (long)(Vector.Length), PtrVector, ev);
If I write the "event" parameter of the function to NULL the execution time is the same in both cards (code show bellow):
GPUDevice.oclCQ.Write<byte>((CL.ComputeBuffer<byte>)Memory.Mem, true, 0, (long)(Vector.Length), PtrVector,null);
The same happend when I execute kernels. I also tested another Opencl Component for .net and work fine on both cards. Why this happend?.
Log in to post a comment.
Hi,
I tested my code in 2 differents GPUs :
- GeForce 9300M GS (OpenCL 1.0)
- GTX 580 (OpenCL 1.1)
When I execute the code show bellow, in 9300M card its more fast than gtx 580:
If I write the "event" parameter of the function to NULL the execution time is the same in both cards (code show bellow):
The same happend when I execute kernels. I also tested another Opencl Component for .net and work fine on both cards. Why this happend?.