Red Hat Work On Improving Glibc Math Performance
Siddhesh Poyarekar of Red Hat has written a blog post about how he's improved the multi-precision number math performance for glibc.
Siddhesh pushed into glibc-2.18 upstream performance improvements of about eight times for the slowest path of the pow function inside glibc. Other functions received significant performance improvements too thanks to improving the multi-precision code-path.
Further math performance improvements for glibc are still planned but with this accomplished work is likely the end of the "low hanging fruit" for gaining great performance improvements out of glibc's existing multiple precision code.
Those interested in learning more of the details of this glibc math performance work, see this Red Hat developer blog post.
Siddhesh pushed into glibc-2.18 upstream performance improvements of about eight times for the slowest path of the pow function inside glibc. Other functions received significant performance improvements too thanks to improving the multi-precision code-path.
Further math performance improvements for glibc are still planned but with this accomplished work is likely the end of the "low hanging fruit" for gaining great performance improvements out of glibc's existing multiple precision code.
Those interested in learning more of the details of this glibc math performance work, see this Red Hat developer blog post.
4 Comments