Since we now know that Navi features ray tracing (at least the PS5 custom variant does), could it be possible that AMD has added AVX to accelerate ray tracing for the platform?

We do know that AVX-512 is great for ray tracing because it has a very good floating point throughput, which is one critical area for ray tracing performance.

While AMD has no existing AVX-512 implementation, could they have feasibly used 2x 256-bit registers to achieve AVX-512 workloads (like how they used 2x 128-bit SIMD registers for 256-bit AVX2) in the PS5 Navi to mix and match the better floating point throughput with massive GPU parallelization to make it possible?