I benchmarked over 130 LLM configs with multiple stacked 16GB "e-waste" GPUs (V100, P100, K80)

🌐 链接: https://www.reddit.com/r/LocalLLaMA/comments/1xymcjb/i_benchmarked_over_130_llm_configs_with_multiple/
🏷️ 分组: Reddit|AI·自托管·VPS
🕒 时间: 2026-11-17 22:49:19 From the LocalLLaMA community on Reddit: I benchmarked over 130 LLM configs with multiple stacked 16GB "e-waste" GPUs (V100, P100…