Thank you for the fast reply and for the correction. How many SIMD groups (possibly belonging to different threadgroups) can be executed in parallel by a single shader core? Even though I made an incorrect assumption, the WWDC session I cited still mentions "1K to 2K concurrent threads per shader core is considered a very good occupancy". It sounds to me like a single shader core is capable of running at least 2K threads (64 SIMD groups) in parallel. But what is the maximum amount?

This fixed the issue for me! Thank you so much!!