Here's an article about JITing x86 to AVX-512 to fuzz 16 VMs per thread:
https://gamozolabs.github.io/fuzzing/2018/10/14/vectorized_e...