Reproducible results across GPUs are theoretically possible, but it would take some effort to implement and be slower. At least the primitive operations (additi

Reproducible results across GPUs are theoretically possible, but it would take some effort to implement and be slower. At least the primitive operations (addition, multiplication, etc.) are there: https://docs.nvidia.com/cuda/floating-point/index.html https://docs.nvidia.com/cuda/floating-point/index.html