GPGPU: V-RAY rendering
With VRAY we can measure with a benchmark two test scenes over GPGPU or GPU acceleration, one for GPUs and another for CPUs, depending on the processor or GPU (OpenCL/CUDA) type you’d like to measure. The way this works is simple, the software will choose the fastest GPGPU API. If there is a CUDA GPU in use then it uses the GPU CUDA pathway. If there is no CUDA enabled device then it reverts to OpenCL (AMD). There is no manual switch that allows us to enforce say OpenCL only.
You will notice that the RTX 2070 is faster than the RTX 2080, we measured this on different platforms as well as retesting all cards. The test results are valid.