Kind2/bench
2022-07-19 20:23:07 -03:00
..
conv_eval.idr Add conv_eval 2022-07-19 20:23:07 -03:00
conv_eval.kind2 Add conv_eval 2022-07-19 20:23:07 -03:00
nat_exp.agda Add Idris2 benchmark 2022-07-14 13:06:28 -03:00
nat_exp.idr Add Idris2 benchmark 2022-07-14 13:06:28 -03:00
nat_exp.kind2 Fix quote issue 2022-07-17 14:40:28 -03:00
nat_exp.lean Add Lean benchmark 2022-07-14 20:23:26 -03:00
nat_exp.v Add Coq benchmark 2022-07-15 15:22:06 -03:00
README.md Polishments 2022-07-17 01:13:40 -03:00

Bench

Benchmarks of Kind2 vs other proof assistants. These benchmarks measure the time each assistant takes to verify a given file. Replicate as follows:

Agda

time agda -i src file.agda

Coq

time coqtop -l file.v -batch -type-in-type

Idris2

time idris2 --check file.idr

Lean

time lean file.lean

Kind2

time kind2 check file.kind2

Kind2-C

kind2 gen-checker file.kind2
hvm compile file.check.hvm
clang file.check.c -O3 -o check
time ./check

Results

nat_exp

Computes 3 ^ 8, using the Nat type, on the type level.

language | time
-------- | --------
Kind2-C  |   0.17 s
Kind2    |   0.58 s
Agda     |  15.55 s
Idris2   |  67.40 s
Lean     |  timeout