Re:Intel C/C++ compilers complete adoption of LLVM
by mshockwave on Thursday August 12, 2021 @01:24PM
(#28158996)
Exactly, many comments here are saying the difference is in LLVM IR ("middle-end") optimizations but IMHO more secrets are hidden in the backend. For example the latency of different instructions (and their operands) which has huge impact on the overall scheduling.
I assumed the icc secret sauce was in the backend: deep knowledge of how each and every single operation is implemented in every uarch.
Maybe they've written their own x86 backend for LLVM, and are using the rest of LLVM for IR->IR transformations and vectorization?