GCC 10 vs. Clang 10 Compiler Performance On AMD Zen 2 + Intel Cascade Lake

[phoronix]

Phoronix: GCC 10 vs. Clang 10 Compiler Performance On AMD Zen 2 + Intel Cascade Lake

We delivered many benchmarks of Clang 10.0 on various CPUs following that updated LLVM compiler stack release earlier this year. With GCC 10 released earlier this month, we have begun our benchmarking of this annual feature release to the GNU Compiler Collection. First up is a look at the GCC 9 vs. GCC 10 vs. LLVM Clang 10 compiler performance on AMD Zen 2 and Intel Cascade Lake systems.

http://www.phoronix.com/vr.php?view=29191

[carewolf]

Would love at some point to see a comparison in -O2 levels or generic tuning, just to compare them as compilers for general binaries.

[CochainComplex]

So far we haven't seen many scenarios where the GCC 10 performance is vastly different from that of GCC 9 with the CPU microarchitecture families tested so far, but in some areas there are minor improvements to be found

maybe more differences with LTO or PGO flags?

[eltomito]

g! C! C! Yeah!!

[hubicka]

With -fPIC/-fpic GCC defaults to -fsemantic-interposition and clang to -fno-semantic-interposition. So when comparing shared libraries (as done few times here) one needs to pick one of the two for both compilers to get meaningful results.

[DanglingPointer]

Perhaps next time throw in ICC as well and see how it compares when on AMD CPUs?

[Michael]

Perhaps next time throw in ICC as well and see how it compares when on AMD CPUs?

I don't have any copy of ICC.

[Guest]

What about kernel compilation benchmarks between GCC and LLVM? Would be interesting to see those (or have they been done already?)