Show Your Support: This site is primarily supported by advertisements. Ads are what have allowed this site to be maintained on a daily basis for the past 18+ years. We do our best to ensure only clean, relevant ads are shown, when any nasty ads are detected, we work to remove them ASAP. If you would like to view the site without ads while still supporting our work, please consider our ad-free Phoronix Premium.
Apple Mac OS X OpenCL Performance vs. Linux
These tests provide interesting metrics comparing the OpenCL GPU performance with NVIDIA hardware under Mac OS X and Ubuntu Linux along with a look at the CPU versus GPU OpenCL performance under Mac OS X. Using OpenCL on the NVIDIA GeForce 9400M was 77% faster than on the dual-core Intel P7350 at 2GHz, which isn't surprising as although this is a relatively GPU, the Open Computing Language is really designed around exploiting the capabilities of a graphics processor. What is interesting though is that Apple's OpenCL GPU implementation was 61% faster than the NVIDIA Linux OpenCL implementation for the JuliaGPU test.
With the MandelGPU test it was actually faster running it on the CPU than the GPU, which may be due to the OpenCL kernel not being well optimized or just the lower-end GPU struggling here, but with the Intel Core 2 Duo its speed was better by 29%. Comparing the NVIDIA OpenCL performance between Apple Mac OS X and Ubuntu Linux the computational speed was much closer this time around, but Apple's code was faster by 2.5%, which was statistically significant based upon our multi-run values.
When running SmallPT GPU with the caustic rendering scene, the NVIDIA OpenCL driver under Linux finally was faster than that of Apple's with a lead of 46%. Running OpenCL on the CPU under Mac OS X had not worked for this OpenCL benchmark.