[SOLVED] CPU usage spikes

Hi, I'm getting regular CPU spikes, running inference with 5 agents (72 vector obervations and 2x3 discrete actions each, no ray or visual detection, v1.0, windows). There are a couple of other things going on, but ML-Agents appears to cause most of the CPU load. Is there a way to mitigate this? Thanks!

EDIT: The CPU is an i7-4770 3.4 GHz, system has 16 GB RAM. I'm not using a GPU for inference. Do I just need faster hardware?

EDIT: The spikes seem to have gone, after I reloaded my scene.