I don't know how SVE works but I thought the point of it was to let implementations pick a larger size than the CPU supports and then get an automatic speedup f

I don't know how SVE works but I thought the point of it was to let implementations pick a larger size than the CPU supports and then get an automatic speedup from better processors with more vector lanes.