Simple performance test for Sharp (js) and Bimg (golang). Both libraries use libvips bindings (a C library for image processing) to provide an API for image processing on their corresponding language.
- go
$ go test -bench=. -benchtime=100x
goos: darwin
goarch: amd64
pkg: example.com/user/image/process
BenchmarkRotate-8 100 345388913 ns/op
BenchmarkResize300x300-8 100 49381323 ns/op
BenchmarkResize500x500-8 100 80474150 ns/op
BenchmarkSharpened-8 100 645960945 ns/op
BenchmarkWatermarker-8 100 148487707 ns/op
BenchmarkConvertPNG-8 100 601702697 ns/op
BenchmarkCrop-8 100 167317632 ns/op
PASS
ok example.com/user/image/process 206.064s
- js
$ npm run bench
> [email protected] bench /Users/dsbrgg/local/image-handler-tests/js
> node indexBenchmark.js
------------------------------------------------------------------------
rotate [100x] x 0.14 ops/sec ±0.00% (0 runs sampled)
rotate [100x] x 0.14 ops/sec ±0.00% (1 run sampled)
rotate [100x] x 0.14 ops/sec ±12.58% (2 runs sampled)
rotate [100x] x 0.14 ops/sec ±2.60% (3 runs sampled)
rotate [100x] x 0.14 ops/sec ±1.50% (4 runs sampled)
rotate [100x] { moe: 0.08418025708037821,
rme: 1.1830296194571508,
sem: 0.0303243001010008,
deviation: 0.06780719639594153,
mean: 7.1156508422,
sample:
[ 7.172829661, 7.032237233, 7.14505701, 7.174456645, 7.053673662 ],
variance: 0.004597815883077786 }
------------------------------------------------------------------------
resize(300, 300) [100x] x 1.36 ops/sec ±0.00% (0 runs sampled)
resize(300, 300) [100x] x 1.53 ops/sec ±0.00% (1 run sampled)
resize(300, 300) [100x] x 1.52 ops/sec ±78.57% (2 runs sampled)
resize(300, 300) [100x] x 1.53 ops/sec ±17.52% (3 runs sampled)
resize(300, 300) [100x] x 1.51 ops/sec ±9.87% (4 runs sampled)
resize(300, 300) [100x] x 1.53 ops/sec ±6.80% (5 runs sampled)
resize(300, 300) [100x] x 1.53 ops/sec ±5.31% (6 runs sampled)
resize(300, 300) [100x] x 1.49 ops/sec ±4.36% (7 runs sampled)
resize(300, 300) [100x] x 1.50 ops/sec ±3.65% (8 runs sampled)
resize(300, 300) [100x] x 1.49 ops/sec ±3.14% (9 runs sampled)
resize(300, 300) [100x] x 1.49 ops/sec ±2.76% (10 runs sampled)
resize(300, 300) [100x] x 1.49 ops/sec ±2.45% (11 runs sampled)
resize(300, 300) [100x] { moe: 0.014791233503279896,
rme: 2.2140836890932643,
sem: 0.006720233304534255,
deviation: 0.023279571044339642,
mean: 0.6680521416666667,
sample:
[ 0.737498661,
0.651598556,
0.657026562,
0.653151676,
0.660571082,
0.653017121,
0.653424262,
0.672298474,
0.668755532,
0.669659025,
0.669142329,
0.67048242 ],
variance: 0.0005419384280084566 }
------------------------------------------------------------------------
resize(500, 500) [100x] x 0.90 ops/sec ±0.00% (0 runs sampled)
resize(500, 500) [100x] x 0.92 ops/sec ±0.00% (1 run sampled)
resize(500, 500) [100x] x 0.94 ops/sec ±12.30% (2 runs sampled)
resize(500, 500) [100x] x 0.95 ops/sec ±5.64% (3 runs sampled)
resize(500, 500) [100x] x 0.96 ops/sec ±3.94% (4 runs sampled)
resize(500, 500) [100x] x 0.89 ops/sec ±3.22% (5 runs sampled)
resize(500, 500) [100x] x 0.96 ops/sec ±3.35% (6 runs sampled)
resize(500, 500) [100x] x 0.94 ops/sec ±2.92% (7 runs sampled)
resize(500, 500) [100x] x 0.97 ops/sec ±2.47% (8 runs sampled)
resize(500, 500) [100x] { moe: 0.0249334920645839,
rme: 2.335126249765324,
sem: 0.0108124423523781,
deviation: 0.0324373270571343,
mean: 1.067757774,
sample:
[ 1.107796532,
1.086559084,
1.058674555,
1.049521691,
1.041295015,
1.127288374,
1.044998038,
1.061139977,
1.0325467 ],
variance: 0.0010521801866114969 }
------------------------------------------------------------------------
sharpen [100x] x 0.04 ops/sec ±0.00% (0 runs sampled)
sharpen [100x] x 0.04 ops/sec ±0.00% (1 run sampled)
sharpen [100x] x 0.04 ops/sec ±8.77% (2 runs sampled)
sharpen [100x] x 0.05 ops/sec ±3.65% (3 runs sampled)
sharpen [100x] x 0.05 ops/sec ±3.70% (4 runs sampled)
sharpen [100x] { moe: 0.7265975326630086,
rme: 3.2282603305484763,
sem: 0.26174262703998874,
deviation: 0.5852743066707894,
mean: 22.507402076199998,
sample:
[ 22.907989553,
23.226350987,
22.554092604,
21.997102006,
21.851475231 ],
variance: 0.3425460140489732 }
------------------------------------------------------------------------
* This API does not exist on `bimg`
modulate [100x] x 0.06 ops/sec ±0.00% (0 runs sampled)
modulate [100x] x 0.06 ops/sec ±0.00% (1 run sampled)
modulate [100x] x 0.05 ops/sec ±5.98% (2 runs sampled)
modulate [100x] x 0.05 ops/sec ±12.75% (3 runs sampled)
modulate [100x] x 0.06 ops/sec ±7.04% (4 runs sampled)
modulate [100x] { moe: 0.8888445566855703,
rme: 4.858305992351004,
sem: 0.32018896134206426,
deviation: 0.715964283205908,
mean: 18.2953597012,
sample:
[ 17.762271093,
17.595748991,
19.291740005,
18.77134861,
18.055689807 ],
variance: 0.5126048548265496 }
------------------------------------------------------------------------
toPNG [100x] x 0.02 ops/sec ±0.00% (0 runs sampled)
toPNG [100x] x 0.02 ops/sec ±0.00% (1 run sampled)
toPNG [100x] x 0.02 ops/sec ±0.85% (2 runs sampled)
toPNG [100x] x 0.02 ops/sec ±0.17% (3 runs sampled)
toPNG [100x] x 0.02 ops/sec ±4.05% (4 runs sampled)
toPNG [100x] { moe: 1.5399598250194424,
rme: 2.8395002459398166,
sem: 0.5547405709724217,
deviation: 1.2404376265713815,
mean: 54.2334809522,
sample:
[ 53.7181583,
53.646667304,
53.681197278,
56.451960925,
53.669420954 ],
variance: 1.5386855054140423 }
------------------------------------------------------------------------
crop [100x] x 0.68 ops/sec ±0.00% (0 runs sampled)
crop [100x] x 0.69 ops/sec ±0.00% (1 run sampled)
crop [100x] x 0.67 ops/sec ±6.64% (2 runs sampled)
crop [100x] x 0.69 ops/sec ±3.70% (3 runs sampled)
crop [100x] x 0.69 ops/sec ±2.08% (4 runs sampled)
crop [100x] x 0.68 ops/sec ±1.45% (5 runs sampled)
crop [100x] x 0.68 ops/sec ±1.11% (6 runs sampled)
crop [100x] x 0.69 ops/sec ±0.94% (7 runs sampled)
crop [100x] { moe: 0.012288153727718505,
rme: 0.8396147976616669,
sem: 0.005195836671339748,
deviation: 0.0146960453769683,
mean: 1.4635465885,
sample:
[ 1.463463409,
1.448240657,
1.4913849049999999,
1.4541187230000001,
1.455145761,
1.469576043,
1.475601868,
1.450841342 ],
variance: 0.00021597374972191132 }
------------------------------------------------------------------------
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent