I feel I just need to run a slightly too large LLM with too much context on a MBP, and it's enough to slow it down irreparably until it suddenly hard resets. Ma

I feel I just need to run a slightly too large LLM with too much context on a MBP, and it's enough to slow it down irreparably until it suddenly hard resets. Maybe the memory pressure it does that at is much higher though compared to Linux?