TechEmpower Benchmarks

[robgraeber]

trafficstars

I'm just curious how come the techempower benchmarks are showing mitol with a ~20% improvement over the default implementation, while the readme is claiming a 3x improvement?

https://www.techempower.com/benchmarks/#section=data-r15&hw=ph&test=plaintext&l=h4s9h9

[Helidium]

Thank you for your question, I have to investigate the reason of the published results.

In the early edition results were much different: https://www.techempower.com/benchmarks/previews/round15/r15p1-vs-r15p2.html