Yes, the technology page calls out that on average it is 3.7X faster with just LLVM and algorithmic improvements. We are trying to push more value into Vitesse

Yes, the technology page calls out that on average it is 3.7X faster with just LLVM and algorithmic improvements. We are trying to push more value into Vitesse X by adding vector processing, and maybe include some of our threading work in the next version.