My 1ms solution (plus a dumb and slightly less dum...
# advent-of-codeb
bjonnh
12/02/2020, 3:33 AMMy 1ms solution (plus a dumb and slightly less dumb): https://github.com/bjonnh/advent_of_code/blob/master/src/main/kotlin/y2020/day01/main.kt
bjonnh
12/02/2020, 3:48 AMIntArrays for sure are fast
bjonnh
12/02/2020, 3:52 AM1ms for the hyperoptimized, 13ms for the intermediary 35ms for the dumb version
bjonnh
12/02/2020, 3:55 AMoh I can even go below that… now I need to use nanos…
bjonnh
12/02/2020, 3:57 AMfirst run is 0.8ms
bjonnh
12/02/2020, 4:13 AMmaking it sorted also makes it much faster… down to 0.3 (but loading time of course is slower)
e
ephemient
12/02/2020, 4:13 AMmodified your code to run part 1 and 2 separately, threw it into my jmh benchmark, this was the result:
b
bjonnh
12/02/2020, 4:14 AMCool!
e
ephemient
12/02/2020, 4:15 AMI'd suggest setting up JMH, or at least make a test harness that can repeatedly run your functions
b
bjonnh
12/02/2020, 4:15 AMyep I need to try that
bjonnh
12/02/2020, 4:15 AMwhat's your machine like?
e
ephemient
12/02/2020, 4:16 AMLinux, i7-8809G
b
bjonnh
12/02/2020, 4:16 AMok seem coherent with what I see then
bjonnh
12/02/2020, 4:34 AMthanks, jmh is setup now
bjonnh
12/02/2020, 4:34 AMseems to be working well
bjonnh
12/02/2020, 4:37 AMrunning case 1 and 2 together: 3ms/op for dumb, 1.6ms/op for intermediary 0.039ms/op for crazyOpt
bjonnh
12/02/2020, 4:38 AMtaking the worst case, the crazyOpt is ~38 times faster… that's decent 😄
bjonnh
12/02/2020, 4:38 AMThanks for your config, there are quite a few things in there I'll steal from you
bjonnh
12/02/2020, 4:41 AM5 Views