benchmark pedersen vs. sha256 hash in key computation scheme

[gballet]

No description provided.

[gballet]

Getting the sample data with:

go test ./trie/utils/ -run - -bench Bench --count 10 | tee pedersen.txt sha.txt

I then edit these two text files to only keep one method (pedersen or sha) and then rename the test to Hash-8.

Results of the comparison:

> benchstat sha.txt pedersen.txt
name    old time/op    new time/op       delta
Hash-8    1.27µs ±26%  359836.63µs ±82%  +28346885.22%  (p=0.000 n=10+10)

name    old alloc/op   new alloc/op      delta
Hash-8     32.0B ± 0%  27594733.9B ± 0%  +86233443.36%  (p=0.000 n=10+8)

name    old allocs/op  new allocs/op     delta
Hash-8      1.00 ± 0%    431965.33 ± 0%  +43196433.33%  (p=0.000 n=10+9)

[gballet]

This branch is in flux after a botched rebase, nonetheless #53 introduces the fix:

name    old time/op    new time/op     delta
Hash-8     930ns ±15%   589948ns ±12%   +63312.09%  (p=0.000 n=10+8)

name    old alloc/op   new alloc/op    delta
Hash-8     32.0B ± 0%  222662.8B ± 0%  +695721.25%  (p=0.000 n=10+10)

name    old allocs/op  new allocs/op   delta
Hash-8      1.00 ± 0%    3098.00 ± 0%  +309700.00%  (p=0.000 n=10+9)

63k-fold slower instead of 28M-fold slower. That's much better. Closing.