I'm not talking about rewrites and GPUs. I'm saying we do not have dynamic recompilation of everything. As in – if we would have ALL binaries that run on the ma

I'm not talking about rewrites and GPUs. I'm saying we do not have dynamic recompilation of everything. As in – if we would have ALL binaries that run on the machine (starting with the kernel) stored in some portable representation like wasm (or not-fully-portable-but-still-reoptimizable like llvm bitcode) and recompiled with optimization for the current exact processor when starting. Only that would solve the "new generation of VLIW-CPU needs very different compiler optimizations to perform, oops all your binaries are for first generation and they are slow now" problem. GPUs do work like this – shaders recompiled all the time – so VLIW was used in GPUs (e.g. TeraScale). But on CPUs we have a world of optimized, "done" binaries.