Are you using OpenCL (AMD) or Cuda/Optix (Nvidia) compute?
With OpenCL there is a slowdown, and tehre is a thread on it already. More details would be helpful to help you
Are you using OpenCL (AMD) or Cuda/Optix (Nvidia) compute?
With OpenCL there is a slowdown, and tehre is a thread on it already. More details would be helpful to help you