And the bulk of the Telum processor is an inference engine, in essence a very large vector processor (although one tailored for AI inference workloads).