Benchmark #
Benchmarks are done with AMD Ryzen 7 PRO 5850U
with a customized image set: totaling 2.6 GB. Around 80% of the images were taken with a Sony A7 camera, with file sizes averaging around 15 MiB. The remaining 20% were smaller images with sizes ranging from 1 MiB to 5 MiB.
[Nova@WebP webp_bench]$ du -ch pics
17M pics/pics/MC
6.1M pics/pics/Wallpapers
2.0M pics/pics/elk
151M pics/pics/TUNA-SFD
32M pics/pics/ThinkPad
322M pics/pics
5.2M pics/exif-orientation-examples
2.6G pics
2.6G total
For more info on benchmark framework, or you’d like to do a benchmark yourself, visit https://github.com/webp-sh/webp_bench.
Overall test results #
For time consumed(shorter is better)
For RAM usage(smaller is better)
Version | RAM Usage(jemalloc) | Time (jemalloc) | RAM Usage(glibc) | Time (glibc) | RAM Usage(tcmalloc) | Time (tcmalloc) |
---|---|---|---|---|---|---|
0.8.0 | 821 | 59 | 884 | 53 | 776 | 54 |
0.8.1 | 534 | 60 | 562 | 53 | 557 | 55 |
0.8.2 | 567 | 60 | 582 | 54 | 565 | 60 |
0.8.3 | 586 | 60 | 559 | 55 | 558 | 68 |
0.8.4 | 569 | 61 | 552 | 60 | 596 | 62 |
0.9.0 | 543 | 94 | 536 | 90 | 543 | 97 |
0.9.1 | 931 | 198 | 893 | 199 | 911 | 203 |
0.9.2 | 1127 | 203 | 1114 | 196 | 1094 | 200 |
0.9.3 | 1165 | 196 | 1043 | 195 | 973 | 194 |
0.9.4 | 1134 | 189 | 1069 | 195 | 1111 | 196 |
0.9.5 | 626 | 61 | 603 | 60 | 565 | 61 |
0.9.6 | 621 | 61 | 510 | 60 | 586 | 61 |
0.9.7 | 644 | 97 | 643 | 95 | 677 | 97 |
0.9.8 | 627 | 97 | 666 | 92 | 625 | 99 |
0.9.9 | 613 | 96 | 652 | 88 | 624 | 102 |
0.9.10 | 638 | 99 | 679 | 100 | 635 | 98 |
Detailed test results on each version #
X axis is time consumed(shorter is better), Y axis is RAM usage(smaller is better).
Version | glibc | jemalloc | tcmalloc |
---|---|---|---|
0.8.0 | |||
0.8.1 | |||
0.8.2 | |||
0.8.3 | |||
0.8.4 | |||
0.9.0 | |||
0.9.1 | |||
0.9.2 | |||
0.9.3 | |||
0.9.4 | |||
0.9.5 | |||
0.9.6 | |||
0.9.7 | |||
0.9.8 | |||
0.9.9 | |||
0.9.10 |
Benchmark on convert size(compression rate) #
Benchmarks are done on version before 0.6.0, as for versions after 0.6.0, our internal benchmark shows about 5 times faster.
8 core #
file_size_range | file_num | src_size | dist_size | total | user | system | cpu | core |
---|---|---|---|---|---|---|---|---|
(10KB,500KB) | 4600 | 1.3G | 310M | 1:44.49 | 905.41s | 9.55s | 875% | 8 |
(500KB,1MB) | 3500 | 2.4G | 361M | 2:04.81 | 1092.50s | 7.98s | 881% | 8 |
(1MB,4MB) | 2000 | 3.8G | 342M | 2:32.64 | 1345.73s | 10.84s | 888% | 8 |
(4MB,32MB) | 500 | 3.6G | 246M | 1:44.53 | 916.91s | 12.03s | 888% | 8 |
1,2,4,8 core #
file_size_range | file_num | src_size | dist_size | 8 | 4 | 2 | 1 |
---|---|---|---|---|---|---|---|
(10KB,500KB) | 4600 | 1.3G | 310M | 1:44.49 | 2:18.49 | 3:36.05 | 5:20.88 |
(500KB,1MB) | 3500 | 2.4G | 361M | 2:04.81 | 2:49.46 | 4:16.41 | 6:28.97 |
(1MB,4MB) | 2000 | 3.8G | 342M | 2:32.64 | 3:26.18 | 5:22.15 | 7:53.45 |
(4MB,32MB) | 500 | 3.6G | 246M | 1:44.53 | 2:21.22 | 3:39.16 | 5:28.65 |