Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Recent single-core SPECint data misleading #4

Open
codebent opened this issue Jan 22, 2020 · 1 comment
Open

Recent single-core SPECint data misleading #4

codebent opened this issue Jan 22, 2020 · 1 comment

Comments

@codebent
Copy link

I know that it's just the way things have progressed in the benchmarking game, but most of the "single-core" speed results use auto-parallelization and do not truly represent single-core performance. The last non autopar'd x86 results were submitted in 2009 and it topped out at 23.1. I would guess that without autopar current systems would be in the ~50 range. So not a 4x increase in performance over 10 years, but closer to 2.

@karlrupp
Copy link
Owner

@codebent thanks for pointing this out. Yes, SPECint is only an approximation to single-core performance and somewhat biased. However, it is a standardized metric that can be compared across a broad range of processors and many people are aware of its shortcomings.

In the end, it's a log-scale plot, so if we overestimate single-core performance by a factor of two, it doesn't change the overall picture :-)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants