I am developing a OpenCL application that will take advantage of hardware acceleration on FPGA cyclone V. The thing is that the compilation process is too slow (at least one hour) and it is not efficient for a big project, Therefore, I am using opencl over x86/x64 intel platform, to assure that the code and functionality is correct without take care for performance. Performance test will be carried out at the end of the test over fpga.
- FPGA SOC cyclone V support OpenCL1.0, is there any version of intel opencl sdk that include development using OpenCL1.0 and not 2.x as the last version of the sdk?
- The flow of development is correct for this particular issue (design first over intel x86/x64 platform without be worried for the impact of performance, and then compile and test the final design for cyclone V fpga)? Do you recommend me another approach for improve the flow of the development in this particular scenario?
I will deeply appreciate your help. Thank you very much.
Thanks for your question. For now this forum is remaining focused on CPU/GPU OpenCL. The best place to go with FPGA OpenCL questions is the Altera OpenCL Forum. Alternately, if you have a mySupport account you can file a service request (SR) and someone will assist you.