I would expect the impact of cpu limits to be different between K8s providers. I have only used memory limits. If I had a pod that tended to be very cpu inten

I would expect the impact of cpu limits to be different between K8s providers. I have only used memory limits. If I had a pod that tended to be very cpu intensive, I would schedule it on it's own node group.