V100

A Practical llama.cpp Multi-GPU Benchmarking Approach: Is 2x V100 16GB Faster Than One 32GB Card?