What is the best hardware concurrency for running inference on CPU?

Firefox AI Runtime now employs multiple threads to accelerate CPU execution for machine learning tasks. Tests revealed significant performance gains with concurrent threads, but excessive thread counts can be counterproductive. To optimize this, the team developed an approach considering physical cores and core types (performance vs. efficiency) instead of just logical cores. This method, implemented in MLUtils.getOptimalCPUConcurrency, aims to better utilize hardware across CPU architectures while preventing resource overcommitment.