Uint32-to-Uint32 Map
This repository contains an implementation of uint32-to-uint32 map which is ~20-50% faster than Go standard map for the same types (see benchmarks below). The code was based of Brent Pedersen's intintmap and the main logic remains intact, with some bug fixes and improvements of the API itself. The map is backed by a single array which interleaves keys and values to improve data locality.
Usage
// Create a new map with capacity of 1024 (resizeable) and 90% desired fill rate
m := intmap.New(1024, 0.90)
// Store a few key/value pairs
m.Store(1, 100)
m.Store(2, 200)
// Load them
v, ok := m.Load(1)
v, ok := m.Load(2)
// Delete keys
m.Delete(1)
m.Delete(2)
Benchmarks
Looking at the benchmarks agains the standard Go map, this map should perform roughly 20-50% better depending on the conditions.
cpu: Intel(R) Core(TM) i7-9700K CPU @ 3.60GHz
BenchmarkStore/intmap-8 14596622 83.30 ns/op 0 B/op 0 allocs/op
BenchmarkStore/stdmap-8 10043176 116.1 ns/op 0 B/op 0 allocs/op
BenchmarkLoad/intmap-0%-8 10720995 108.1 ns/op 0 B/op 0 allocs/op
BenchmarkLoad/stdmap-0%-8 10443445 110.4 ns/op 0 B/op 0 allocs/op
BenchmarkLoad/intmap-10%-8 10510770 105.3 ns/op 0 B/op 0 allocs/op
BenchmarkLoad/stdmap-10%-8 10319277 109.3 ns/op 0 B/op 0 allocs/op
BenchmarkLoad/intmap-50%-8 12594564 89.33 ns/op 0 B/op 0 allocs/op
BenchmarkLoad/stdmap-50%-8 10215164 112.6 ns/op 0 B/op 0 allocs/op
BenchmarkLoad/intmap-90%-8 19359448 64.39 ns/op 0 B/op 0 allocs/op
BenchmarkLoad/stdmap-90%-8 10160139 111.7 ns/op 0 B/op 0 allocs/op
BenchmarkLoad/intmap-100%-8 20707099 62.17 ns/op 0 B/op 0 allocs/op
BenchmarkLoad/stdmap-100%-8 9713601 110.4 ns/op 0 B/op 0 allocs/op
