Hi I was wondering if Blender supports double precision floating point processing (FP64), as the Titan and now the Titan Black have that ability. From what I’ve seen in the current benchmarks, however, the Titan is not doing that much better than a GTX 780. Any comments on this are appreciated! Thanks.
Only Cycles is CUDA-accelerated in Blender and it does not need double-precision floats. The only thing the Titan has going for it is the larger RAM.
Ah… thanks for the clarification! 