artpar's attempt

[artpar]

• parallel execution
• vector for mix/max/sum
• memory mapped files
• cached parsing for numbers

The numbers which I am seeing on my machine

[base]                   183.28s user 5.63s system 100% cpu 3:08.77 total
[memory mapped+parallel] 167.33s user 5.04s system 630% cpu 27.321 total
[above+vector]           157.12s user 3.43s system 670% cpu 23.958 total

[lobaorn]

Shamelessly sharing this idea for JVM/GC tuning in another PR/discussion? https://github.com/gunnarmorling/1brc/pull/15#issuecomment-1875495420

[artpar]

@lobaorn it makes the runtime worse for my implementation.

[gunnarmorling]

Could you please run test.sh artpar and make sure all the tests pass? Thanks!

[artpar]

Done. Thanks for taking the time to organize this @gunnarmorling

[gunnarmorling]

00:16.196, nice. Thx for participating!