News for hackers. Stuff that matters.
Re:Sorting a Billion Numbers with Julia by en4bz on Monday March 21, 2016 @07:01PM (#11332905)
I was initially skeptical of the 2.5s for the SUM base case but after a bit of experimenting I concluded the following. Note this is for the SUM base case only. The whole file is loaded into memory for my tests.

MMAP'd from HDD 1st run - 47s

MMAP'd from HDD 2nd run - 1.35s (OS caches pages in memory)

MMAP'd from NVMe 1st run - 6.5s (OS Caches dropped)

MMAp'd from NVMe 2nd run 1.35s (OS cached again)

There are no comments on this comment yet.