They already came close in a ton of benchmarks and were faster in others. The biggest issue was AVX. They could run AVX2, but at (effectively) 1/2 clockspeed du

They already came close in a ton of benchmarks and were faster in others. The biggest issue was AVX. They could run AVX2, but at (effectively) 1/2 clockspeed due to their implementation being 128-bits wide instead of 256-bits like Intel's (which downclocked, but was still faster for non-mixed loads). AMD now has 256-bit AVX2 units, but unlike Intel, they don't need to downclock due to 7nm TSMC's lower power requirements compared to Intel's 14nm process. This should also affect 128-bit AVX instructions. It should be possible to reorder and push 2 through the pipeline at the same time in a lot of circumstances.