3 ms·
Odd to not see a mention of LINPACK in that paper, when it has reliably produced temperatures exceeding all other stress-testing programs I've tried, as well as
by userbinator 6y ago
Odd to not see a mention of LINPACK in that paper, when it has reliably produced temperatures exceeding all other stress-testing programs I've tried, as well as uncovered instabilities that others couldn't find. It also turns out to be a very realistic type of benchmark (solving dense systems of linear equations). The paper doesn't predate its use, so I guess the authors just haven't spent much time on overclocking forums...