Benchmarks should've included random ASCII. I understand they hinted at that with labeling their benchmark synthetic, but it'd give a idea of the impact of bran

Benchmarks should've included random ASCII. I understand they hinted at that with labeling their benchmark synthetic, but it'd give a idea of the impact of branch mispredicts