Correct, though generally you’d have region constraints on your workload. We do see a lot of value in going cross-region for expensive tasks that aren’t latency or data transfer sensitive, such as running machine learning model training in cheaper regions.