Thank you! Now it makes sense to me.
On Monday, March 27, 2023 at 5:50:02 PM UTC+3 Keith Randall wrote:
> Key sizes 4 and 8 have special case hashing code.
> They are intended for int32, int64, and pointers, but your [4]byte and
> [8]byte take advantage of it as well.
>
> On Sunday, March 26, 2023 at 3:05:52 AM UTC-7 Amit Lavon wrote:
>
>> Hi gophers,
>>
>> Some code I am writing uses byte-arrays ([X]byte) as keys in a map. I
>> benchmarked the performance of map operations using different X's and found
>> that 4 and 8 are about twice as fast compared to 5, 6, 7 (see below).
>>
>> Can someone explain this phenomenon?
>> I'd like to learn about it so I can take it in consideration when
>> choosing key types.
>>
>>
>> Code:
>>
>> type byteArray interface {
>> [4]byte | [5]byte | [6]byte | [7]byte | [8]byte
>> }
>>
>> func benchmarkMaps[T byteArray](b *testing.B) {
>> m := map[T]int{}
>> var t T
>> for i := 0; i < b.N; i++ {
>> m[t]++
>> }
>> }
>>
>> func BenchmarkMaps(b *testing.B) {
>> b.Run("4", benchmarkMaps[[4]byte])
>> b.Run("5", benchmarkMaps[[5]byte])
>> b.Run("6", benchmarkMaps[[6]byte])
>> b.Run("7", benchmarkMaps[[7]byte])
>> b.Run("8", benchmarkMaps[[8]byte])
>> }
>>
>> Results:
>>
>> 4: 17.01 ns/op
>> 5: 36.02 ns/op
>> 6: 38.24 ns/op
>> 7: 29.92 ns/op
>> 8: 16.58 ns/op
>>
>> go1.20
>>
>
--
You received this message because you are subscribed to the Google Groups
"golang-nuts" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to golang-nuts+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/golang-nuts/93df68a5-3189-4ac5-b25f-6d5d4bd2e125n%40googlegroups.com.
