Don't have Telegram yet? Try it now!
https://hgpu.org/?p=10687
Characterizing the Challenges and Evaluating the Efficacy of a CUDA-to-OpenCL Translator