On Sat, 6 Apr 2024 at 03:34, David Rowley <dgrowle...@gmail.com> wrote: > Does anyone else want to try the attached script on the v5 patch to > see if their numbers are better?
On my machine (i9-10900X, in Ubuntu 22.04 on WSL on Windows) v5 consistently beats master by ~0.25 seconds: master: Run 100 100 5000000: 1.948975205 Run 1024 10240 200000: 3.039986587 Run 1024 1048576 2000: 2.444176276 Run 1048576 1048576 1000: 2.475328596 v5: Run 100 100 5000000: 1.997170909 Run 1024 10240 200000: 3.057802598 Run 1024 1048576 2000: 2.199449857 Run 1048576 1048576 1000: 2.210328762 The first two runs are pretty much equal, and I ran your script a few more times and this seems like just random variance (sometimes v5 wins those, sometimes master does always quite close to each other). But the last two runs v5 consistently wins. Weird that on your machines you don't see a difference. Are you sure you didn't make a silly mistake, like not restarting postgres or something?