Healthy Competition With GCC 15 vs. LLVM Clang 20 Performance On AMD Zen 5

Written by Michael Larabel in Software on 11 February 2025 at 01:03 PM EST. Page 5 of 5. 10 Comments.

The Clang-built binaries of the uvg266 H.266 encoder had a slight advantage over GCC on this AMD Zen 5 Linux server.

👁 x265 benchmark with settings of Video Input: Bosphorus 4K. Clang 19.1.7 was the fastest.

👁 Coremark benchmark with settings of CoreMark Size 666, Iterations Per Second. GCC 15.0.1 Git 31 Jan was the fastest.

👁 7-Zip Compression benchmark with settings of Test: Compression Rating. GCC 15.0.1 Git 31 Jan was the fastest.

👁 7-Zip Compression benchmark with settings of Test: Decompression Rating. GCC 15.0.1 Git 31 Jan was the fastest.

👁 Stargate Digital Audio Workstation benchmark with settings of Sample Rate: 96000, Buffer Size: 512. Clang 19.1.7 was the fastest.

👁 Stargate Digital Audio Workstation benchmark with settings of Sample Rate: 192000, Buffer Size: 512. Clang 19.1.7 was the fastest.

👁 C-Ray benchmark with settings of Resolution: 5K, Rays Per Pixel: 16. Clang 20.1 Git was the fastest.

👁 LAME MP3 Encoding benchmark with settings of WAV To MP3. GCC 15.0.1 Git 31 Jan was the fastest.

👁 Liquid-DSP benchmark with settings of Threads: 1, Buffer Length: 256, Filter Length: 512. Clang 19.1.7 was the fastest.

👁 Liquid-DSP benchmark with settings of Threads: 64, Buffer Length: 256, Filter Length: 512. Clang 19.1.7 was the fastest.

👁 Llama.cpp benchmark with settings of Backend: CPU BLAS, Model: Llama-3.1-Tulu-3-8B-Q8_0, Test: Text Generation 128. GCC 14.2 was the fastest.

👁 Llama.cpp benchmark with settings of Backend: CPU BLAS, Model: Llama-3.1-Tulu-3-8B-Q8_0, Test: Prompt Processing 1024. GCC 14.2 was the fastest.

👁 Llama.cpp benchmark with settings of Backend: CPU BLAS, Model: Mistral-7B-Instruct-v0.3-Q8_0, Test: Text Generation 128. GCC 14.2 was the fastest.

There remains some very healthy competition between the GCC and LLVM Clang compilers on Linux x86_64 with their newest stable and in-development releases. Those wanting to dig through even more raw benchmark data can find all of the tests here.

👁 Geometric Mean Of All Test Results benchmark with settings of Result Composite, LLVM Clang 20 Benchmarks On AMD EPYC Zen 5. GCC 15.0.1 Git 31 Jan was the fastest.

On a geometric mean basis when counting all of the dozens of C/C++ workloads that successfully ran on all four tested compiler versions, GCC 14 and GCC 15 retain a very ever small victory over LLVM Clang 19/20 on this AMD EPYC 9005 series Zen 5 server. This is similar to what we have seen on other Intel/AMD hardware in recent times of both GCC and Clang delivering very healthy competition for these open-source compilers. With some codebases/workloads there can be strong advantages at time for one compiler over the other, but at a high level the GCC and Clang compiler performance is extremely tight with recent versions and on modern x86_64 hardware.

10 Comments

If you enjoyed this article consider joining Phoronix Premium to view this site ad-free, multi-page articles on a single page, and other benefits. PayPal or Stripe tips are also graciously accepted. Thanks for your support.

Page: 1 2 3 4 5

URL: https://www.phoronix.com/review/clang20-gcc15-amd-znver5/5

⇱ Healthy Competition With GCC 15 vs. LLVM Clang 20 Performance On AMD Zen 5 - Phoronix

Healthy Competition With GCC 15 vs. LLVM Clang 20 Performance On AMD Zen 5