3 ms·
Benchmarking really does need repeated inputs and outputs. Otherwise you can't make a meaningful comparison between the systems that ran the test. You can use
by toast0 5y ago
Benchmarking really does need repeated inputs and outputs. Otherwise you can't make a meaningful comparison between the systems that ran the test.
You can use these distributed computing projects for burn-in testing though. When you just want to run your computer hard to make sure it can manage N hours at 100%, and don't necessarily need to know how it compares to others, distributed computing can work (although, there's some risk of wrong results and therefore more work for the organizers if you're testing your system in extreme conditions, sometimes those lead to miscalculations)