I am a new to Opencl. I download example compile and run FluidSimulation and got "Device does not support cl_amd_fp64 extension!"
My Hardware AMD FX8120, GPU 6450 Window 7 Ultimate.
What should i do to run this sample code?
IF the sample says it does not suport cl_amd_fp64, that means double precision is not supported on the GPU. You can still run it on your CPU.
I want to find out why It does't support? I am doing research, I need floating point on GPU. If someone know the answer please help.
The device does support floating point operations. But it does not support double precision arithmetic. There are many samples that take floats as arguments.(NBody for example)
Nguyen Viet Trung
Please see the specifications on the product page: AMD Radeon™ HD 6450 Graphics. If double precision were supported there would be a line indicating how many "GFLOPS Double Precision compute power." If the compute power quoted doesn't specifically mention double precision, it should be interpreted as single precision.
As far as I can tell, double precision is generally supported for Radeon HD x9xx graphics cards, may or may not be supported for x8xx and x7xx, and is probably not supported otherwise (e.g., x6xx, x5xx, and x4xx).
I will buy latest AMD GPU card.
Retrieving data ...