News for hackers. Stuff that matters.

A comment on Silent Data Corruptions at Scale (2021). Parent comment

Re:Silent Data Corruptions at Scale (2021) by jorticka on Tuesday February 20, 2024 @01:19AM (#39438531)
Like all stress tests, linpack will find some errors, but not all.

I had memory stability issues which would immediatly show under Prime95 (less than 1 minute) but pass hours of Linpack.

[ Parent | on HN ]
Re:Silent Data Corruptions at Scale (2021) by sirlancer on Tuesday February 20, 2024 @06:40AM (#39440129)
Prime95 is my gold standard for CPU and memory testing. Everything from desktops to HPC and clustered filesystems get a 24 hour “blend” of tests. If that passes without any instability or bit flips then we’re ready for production.
[ Parent | on HN ]
Re:Silent Data Corruptions at Scale (2021) by c0l0 on Tuesday February 20, 2024 @07:06AM (#39440275)
In my experience, LINPACK (at least the Intel MKL on GenuineIntel combination) is both quicker and more thorough in finding setups that are not actually stable/reliable.
[ Parent | on HN ]