4 ms·
Running 40 AI Agents in Parallel on Mac Studio M3 Ultra
- ekropotin 1y agoInteresting experiment! However, I think it lacks the most interesting information - which is a latency in tokens/sec and how it decays with increasing of parallelism. I also not exactly got what it has to do with AI Agents in particular.