News for hackers. Stuff that matters.
Re:tolower() in bulk at speed by dzaima on Tuesday June 28, 2022 @04:47PM (#31912809)
Between a jump table and a scalar loop, the loop will screw with the branch predictor more.

A jump table has the potential to be faster by doing SIMD, at the cost of taking more icache.

There are no comments on this comment yet.