Unfortunately none of the hardware used for testing supports FP16 arithmetic. Between Intel and AMD, the only platform that supports AVX512-FP16 is currently Sapphire Rapids.

Alderlake supports AVX512-FP16. Still only 9.6x faster than div. Most likely reciprocal is just too slow.

I tried a similar approach with 32-bit FP before, and the problem here is that fast conversion is only fast in the sense of latency. Throughput-wise, it takes 2 uops instead of one, so in the end, a plain float<->int conversion wins.