Open AndrewReitz opened 3 years ago
RandomX is useless on GPUs anyway (note it's only making one thread on such a powerful card)
Compile CPU-only version and separate CUDA version and run them separately but parallel, then they can do different algos (if on MoneroOcean or other multi-mining supported pool).
Or if you only want RandomX based algos, forget using the GPU it's a headache and also horrible hash-per-watt (as designed).
After a few minutes of running the cuda mining stops after the error
Output from the console here. Let me know if there is any more information I can provide.