Benchmarks to compare the performance of wasm compilers
make native wasmedge wasmtime wazero count=10 bench='BenchmarkJoinPath/#00' test=go
make native wasmedge wasmtime wazero count=10 bench='BenchmarkJoinPath/#00' test=tinygo$ benchstat -ignore=goarch,goos bench.tinygo.*
│ bench.tinygo.native │ bench.tinygo.node │ bench.tinygo.wasmedge │ bench.tinygo.wasmtime │ bench.tinygo.wazero │
│ sec/op │ sec/op vs base │ sec/op vs base │ sec/op vs base │ sec/op vs base │
JoinPath/#00 126.85n ± 1% 47.46n ± 1% -62.58% (p=0.000 n=10) 46.40n ± 1% -63.42% (p=0.000 n=10) 74.80n ± 2% -41.03% (p=0.000 n=10) 143.05n ± 1% +12.77% (p=0.000 n=10)
$ benchstat -ignore=goarch,goos bench.go.*
pkg: github.com/stealthrocket/wasm-benchmarks/go
│ bench.go.native │ bench.go.node │ bench.go.wasmedge │ bench.go.wasmtime │ bench.go.wazero │
│ sec/op │ sec/op vs base │ sec/op vs base │ sec/op vs base │ sec/op vs base │
JoinPath/#00 15.99n ± ∞ ¹ 96.75n ± 1% ~ (p=0.182 n=1+10) 80.98n ± ∞ ¹ ~ (p=1.000 n=1) ² 445.00n ± ∞ ¹ ~ (p=1.000 n=1) ² 254.50n ± ∞ ¹ ~ (p=1.000 n=1) ²
$ benchstat -ignore=goarch,goos bench.c.*
│ bench.c.native │ bench.c.node │ bench.c.wasmedge │ bench.c.wasmtime │ bench.c.wazero │
│ sec/op │ sec/op vs base │ sec/op vs base │ sec/op vs base │ sec/op vs base │
JoinPath/#00 24.53n ± 0% 24.29n ± 0% -0.96% (p=0.001 n=10) 26.15n ± 2% +6.63% (p=0.000 n=10) 48.84n ± 1% +99.14% (p=0.000 n=10) 73.01n ± 1% +197.72% (p=0.000 n=10)