Avoiding instruction cache misses21 June 2019Modern processors are quite complex, with many parts having the potential to become a bottleneck. It is relatively easy to reason about the performance of short pieces of code, especially if memory effects are kept to a minimum. Both static analysis tools like LLVM MCA and microbenchmarks can provide a lot of information in such cases. However, the beha
{{#tags}}- {{label}}
{{/tags}}