A simple benchmarking lib in Zig
Test [0/2] test "Debug benchmark"...
Benchmark Iterations Min(ns) Max(ns) Variance Mean(ns)
----------------------------------------------------------------------------
sum_slice(block=16) 100000 90 2690 243 107
sum_slice(block=32) 100000 170 1760 338 190
sum_slice(block=64) 100000 320 2340 476 352
sum_slice(block=128) 100000 630 2290 862 678
sum_slice(block=256) 100000 1270 3170 2402 1336
sum_slice(block=512) 100000 2550 8490 4835 2651
sum_reader(block=16) 100000 990 2640 1592 1039
sum_reader(block=32) 100000 1930 3890 3292 2012
sum_reader(block=64) 100000 3830 6250 6806 3962
sum_reader(block=128) 63673 7660 12830 15703 7852
sum_reader(block=256) 31967 15360 22190 31847 15641
sum_reader(block=512) 16031 30800 34690 59444 31191
Test [1/2] test "Debug benchmark generics"...
Benchmark Iterations Min(ns) Max(ns) Variance Mean(ns)
----------------------------------------------------------------------------
sum_vectors(vec4f16) 100000 2730 13390 3620 2775
sum_vectors(vec4f32) 100000 1289 5800 1277 1296
sum_vectors(vec4f64) 100000 1389 6870 1358 1400
sum_vectors(vec8f16) 100000 4400 9680 4613 4479
sum_vectors(vec8f32) 100000 1389 5180 1231 1400
sum_vectors(vec8f64) 100000 1390 6170 2260 1457
sum_vectors(vec16f16) 61088 8090 13980 15455 8184
sum_vectors(vec16f32) 100000 1399 4560 2069 1441
sum_vectors(vec16f64) 100000 1440 6080 1664 1475
All 2 tests passed.
Test [0/2] test "ReleaseSafe benchmark"...
Benchmark Iterations Min(ns) Max(ns) Variance Mean(ns)
----------------------------------------------------------------------------
sum_slice(block=16) 100000 9 3550 164 18
sum_slice(block=32) 100000 9 940 22 18
sum_slice(block=64) 100000 49 1530 66 52
sum_slice(block=128) 100000 89 1280 102 92
sum_slice(block=256) 100000 169 1690 210 171
sum_slice(block=512) 100000 319 5530 724 329
sum_reader(block=16) 100000 60 2840 180 69
sum_reader(block=32) 100000 110 3059 288 121
sum_reader(block=64) 100000 209 2810 323 224
sum_reader(block=128) 100000 400 1780 387 431
sum_reader(block=256) 100000 790 2220 681 843
sum_reader(block=512) 100000 1550 4300 3805 1669
Test [1/2] test "ReleaseSafe benchmark generics"...
Benchmark Iterations Min(ns) Max(ns) Variance Mean(ns)
----------------------------------------------------------------------------
sum_vectors(vec4f16) 100000 1269 3790 1799 1283
sum_vectors(vec4f32) 100000 319 1680 300 328
sum_vectors(vec4f64) 100000 319 1860 355 329
sum_vectors(vec8f16) 100000 2399 5010 5014 2420
sum_vectors(vec8f32) 100000 319 2660 641 329
sum_vectors(vec8f64) 100000 319 7740 1019 330
sum_vectors(vec16f16) 100000 4599 9970 22580 4636
sum_vectors(vec16f32) 100000 319 4310 1231 330
sum_vectors(vec16f64) 100000 429 4070 1783 439
All 2 tests passed.
Test [0/2] test "ReleaseFast benchmark"...
Benchmark Iterations Min(ns) Max(ns) Variance Mean(ns)
----------------------------------------------------------------------------
sum_slice(block=16) 100000 19 2840 128 21
sum_slice(block=32) 100000 19 1600 78 20
sum_slice(block=64) 100000 19 1970 74 21
sum_slice(block=128) 100000 19 1530 68 21
sum_slice(block=256) 100000 39 1250 74 44
sum_slice(block=512) 100000 59 1150 85 68
sum_reader(block=16) 100000 19 1170 21 20
sum_reader(block=32) 100000 19 1650 74 20
sum_reader(block=64) 100000 19 1250 34 20
sum_reader(block=128) 100000 19 1240 32 20
sum_reader(block=256) 100000 39 2180 177 44
sum_reader(block=512) 100000 59 2470 148 68
Test [1/2] test "ReleaseFast benchmark generics"...
Benchmark Iterations Min(ns) Max(ns) Variance Mean(ns)
----------------------------------------------------------------------------
sum_vectors(vec4f16) 100000 1259 8590 1678 1284
sum_vectors(vec4f32) 100000 319 1440 279 327
sum_vectors(vec4f64) 100000 319 1760 303 327
sum_vectors(vec8f16) 100000 2399 5260 1861 2417
sum_vectors(vec8f32) 100000 319 2080 434 327
sum_vectors(vec8f64) 100000 319 1710 329 328
sum_vectors(vec16f16) 100000 4599 9010 3883 4634
sum_vectors(vec16f32) 100000 319 2800 356 329
sum_vectors(vec16f64) 100000 429 1750 404 436
All 2 tests passed.
Test [0/2] test "ReleaseSmall benchmark"...
Benchmark Iterations Min(ns) Max(ns) Variance Mean(ns)
----------------------------------------------------------------------------
sum_slice(block=16) 100000 19 2760 247 27
sum_slice(block=32) 100000 29 5090 363 37
sum_slice(block=64) 100000 50 2640 177 63
sum_slice(block=128) 100000 90 1830 157 102
sum_slice(block=256) 100000 169 5860 733 201
sum_slice(block=512) 100000 330 3690 1560 365
sum_reader(block=16) 100000 219 1430 276 226
sum_reader(block=32) 100000 420 1870 460 432
sum_reader(block=64) 100000 819 2690 770 837
sum_reader(block=128) 100000 1629 5390 1696 1649
sum_reader(block=256) 100000 3240 9080 3240 3274
sum_reader(block=512) 76638 6469 9780 5302 6524
Test [1/2] test "ReleaseSmall benchmark generics"...
Benchmark Iterations Min(ns) Max(ns) Variance Mean(ns)
----------------------------------------------------------------------------
sum_vectors(vec4f16) 100000 4859 16710 5250 4902
sum_vectors(vec4f32) 100000 319 1650 326 328
sum_vectors(vec4f64) 100000 319 1470 295 327
sum_vectors(vec8f16) 100000 3980 9070 3382 4254
sum_vectors(vec8f32) 100000 319 3740 459 328
sum_vectors(vec8f64) 100000 319 4100 534 330
sum_vectors(vec16f16) 79800 6219 15130 10000 6265
sum_vectors(vec16f32) 100000 319 3340 455 330
sum_vectors(vec16f64) 100000 429 2020 454 438
All 2 tests passed.