HTML Preparation code:
Script Preparation code:
Tests:

Historical Performance Data aggregated from 6 community runs

Desktop Results:
Chrome 6 runs
(CPU) 10
41.7M/s
(CPU) 11
41.5M/s
(CPU) 9
41.2M/s
(CPU) 15
40.3M/s
(CPU) 17
39.7M/s
(WebGL) 14
39.6M/s
(WebGL) 13
39.4M/s
(CPU) 13
39.3M/s
(WebGL) 11
38.3M/s
(WebGL) 15
37.9M/s
(WebGL) 12
37.9M/s
(WebGL) 17
37.9M/s
(CPU) 18
37.7M/s
(WebGL) 16
37.6M/s
(CPU) 14
37.5M/s
(CPU) 16
37.3M/s
(WebGL) 18
37.1M/s
(CPU) 12
36.8M/s
(WebGL) 9
34.8M/s
(WebGL) 10
33.6M/s
(CPU) AddGatherReshape_channelMultiplier_1
357K/s
(WebGL) AddGatherReshape_channelMultiplier_1
244K/s
Async wait for init (WebGL)
45K/s
Async wait for init (CPU)
43K/s
Async wait for init (WASM)
34K/s
(WASM) AddGatherReshape_channelMultiplier_1
20K/s
(WebGL) AddGatherReshape_channelMultiplier_2
2K/s
(WebGL) AddGatherReshape_channelMultiplier_4
1K/s
(WebGL) AddGatherReshape_channelMultiplier_8
1K/s
(WebGL) SplitReshapeGatherConcat_channelMultiplier_1
323/s
(WebGL) SplitReshapeGatherConcat_channelMultiplier_2
299/s
(WebGL) SplitReshapeGatherConcat_channelMultiplier_4
255/s
(WebGL) SplitReshapeGatherConcat_channelMultiplier_8
211/s
(CPU) AddGatherReshape_channelMultiplier_2
23/s
(WASM) SplitReshapeGatherConcat_channelMultiplier_1
21/s
(WASM) SplitReshapeGatherConcat_channelMultiplier_2
15/s
(CPU) AddGatherReshape_channelMultiplier_4
15/s
(CPU) SplitReshapeGatherConcat_channelMultiplier_1
10/s
(WASM) SplitReshapeGatherConcat_channelMultiplier_4
10/s
(CPU) AddGatherReshape_channelMultiplier_8
9/s
(CPU) SplitReshapeGatherConcat_channelMultiplier_2
9/s
(WASM) AddGatherReshape_channelMultiplier_2
8/s
(CPU) SplitReshapeGatherConcat_channelMultiplier_4
8/s
(WASM) AddGatherReshape_channelMultiplier_4
7/s
(WASM) SplitReshapeGatherConcat_channelMultiplier_8
6/s
(CPU) SplitReshapeGatherConcat_channelMultiplier_8
6/s
(WASM) AddGatherReshape_channelMultiplier_8
5/s

Fastest: (CPU) 10 (9058825.9x faster)

View exact numbers
  • (CPU) 10: 41,693,328 ops/sec
  • (CPU) 11: 41,466,448 ops/sec
  • (CPU) 9: 41,226,188 ops/sec
  • (CPU) 15: 40,273,576 ops/sec
  • (CPU) 17: 39,734,052 ops/sec
  • (WebGL) 14: 39,608,136 ops/sec
  • (WebGL) 13: 39,387,016 ops/sec
  • (CPU) 13: 39,336,876 ops/sec
  • (WebGL) 11: 38,262,392 ops/sec
  • (WebGL) 15: 37,933,188 ops/sec
  • (WebGL) 12: 37,930,124 ops/sec
  • (WebGL) 17: 37,880,892 ops/sec
  • (CPU) 18: 37,667,920 ops/sec
  • (WebGL) 16: 37,637,084 ops/sec
  • (CPU) 14: 37,539,384 ops/sec
  • (CPU) 16: 37,254,976 ops/sec
  • (WebGL) 18: 37,110,672 ops/sec
  • (CPU) 12: 36,829,348 ops/sec
  • (WebGL) 9: 34,832,704 ops/sec
  • (WebGL) 10: 33,564,128 ops/sec
  • (CPU) AddGatherReshape_channelMultiplier_1: 357,424 ops/sec
  • (WebGL) AddGatherReshape_channelMultiplier_1: 244,342 ops/sec
  • Async wait for init (WebGL): 44,925 ops/sec
  • Async wait for init (CPU): 42,682 ops/sec
  • Async wait for init (WASM): 34,042 ops/sec
  • (WASM) AddGatherReshape_channelMultiplier_1: 19,670 ops/sec
  • (WebGL) AddGatherReshape_channelMultiplier_2: 1,626 ops/sec
  • (WebGL) AddGatherReshape_channelMultiplier_4: 1,377 ops/sec
  • (WebGL) AddGatherReshape_channelMultiplier_8: 1,153 ops/sec
  • (WebGL) SplitReshapeGatherConcat_channelMultiplier_1: 323 ops/sec
  • (WebGL) SplitReshapeGatherConcat_channelMultiplier_2: 299 ops/sec
  • (WebGL) SplitReshapeGatherConcat_channelMultiplier_4: 255 ops/sec
  • (WebGL) SplitReshapeGatherConcat_channelMultiplier_8: 211 ops/sec
  • (CPU) AddGatherReshape_channelMultiplier_2: 23 ops/sec
  • (WASM) SplitReshapeGatherConcat_channelMultiplier_1: 21 ops/sec
  • (WASM) SplitReshapeGatherConcat_channelMultiplier_2: 15 ops/sec
  • (CPU) AddGatherReshape_channelMultiplier_4: 15 ops/sec
  • (CPU) SplitReshapeGatherConcat_channelMultiplier_1: 10 ops/sec
  • (WASM) SplitReshapeGatherConcat_channelMultiplier_4: 10 ops/sec
  • (CPU) AddGatherReshape_channelMultiplier_8: 9 ops/sec
  • (CPU) SplitReshapeGatherConcat_channelMultiplier_2: 9 ops/sec
  • (WASM) AddGatherReshape_channelMultiplier_2: 8 ops/sec
  • (CPU) SplitReshapeGatherConcat_channelMultiplier_4: 8 ops/sec
  • (WASM) AddGatherReshape_channelMultiplier_4: 7 ops/sec
  • (WASM) SplitReshapeGatherConcat_channelMultiplier_8: 6 ops/sec
  • (CPU) SplitReshapeGatherConcat_channelMultiplier_8: 6 ops/sec
  • (WASM) AddGatherReshape_channelMultiplier_8: 5 ops/sec

Results show median performance across all community submissions. Run this benchmark yourself to verify on your device.

Rendered benchmark preparation results:

Suite status: <idle, ready to run>

Previous results
  • Test case name Result
    Async wait for init (WASM)
    (WASM) AddGatherReshape_channelMultiplier_1
    (WASM) SplitReshapeGatherConcat_channelMultiplier_1
    (WASM) AddGatherReshape_channelMultiplier_2
    (WASM) SplitReshapeGatherConcat_channelMultiplier_2
    (WASM) AddGatherReshape_channelMultiplier_4
    (WASM) SplitReshapeGatherConcat_channelMultiplier_4
    (WASM) AddGatherReshape_channelMultiplier_8
    (WASM) SplitReshapeGatherConcat_channelMultiplier_8
    Async wait for init (WebGL)
    (WebGL) AddGatherReshape_channelMultiplier_1
    (WebGL) SplitReshapeGatherConcat_channelMultiplier_1
    (WebGL) AddGatherReshape_channelMultiplier_2
    (WebGL) SplitReshapeGatherConcat_channelMultiplier_2
    (WebGL) AddGatherReshape_channelMultiplier_4
    (WebGL) SplitReshapeGatherConcat_channelMultiplier_4
    (WebGL) AddGatherReshape_channelMultiplier_8
    (WebGL) SplitReshapeGatherConcat_channelMultiplier_8
    (WebGL) 9
    (WebGL) 10
    (WebGL) 11
    (WebGL) 12
    (WebGL) 13
    (WebGL) 14
    (WebGL) 15
    (WebGL) 16
    (WebGL) 17
    (WebGL) 18
    Async wait for init (CPU)
    (CPU) AddGatherReshape_channelMultiplier_1
    (CPU) SplitReshapeGatherConcat_channelMultiplier_1
    (CPU) AddGatherReshape_channelMultiplier_2
    (CPU) SplitReshapeGatherConcat_channelMultiplier_2
    (CPU) AddGatherReshape_channelMultiplier_4
    (CPU) SplitReshapeGatherConcat_channelMultiplier_4
    (CPU) AddGatherReshape_channelMultiplier_8
    (CPU) SplitReshapeGatherConcat_channelMultiplier_8
    (CPU) 9
    (CPU) 10
    (CPU) 11
    (CPU) 12
    (CPU) 13
    (CPU) 14
    (CPU) 15
    (CPU) 16
    (CPU) 17
    (CPU) 18

    Fastest: N/A

    Slowest: N/A

Latest run results:
Run details: (Test run date: one year ago)
Mozilla/5.0 (X11; CrOS x86_64 14541.0.0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/137.0.0.0 Safari/537.36
Chrome 137 on Chrome OS 14541.0.0
View result in a separate tab
(CPU) 14
44.3M/s
(WebGL) 13
44.1M/s
(CPU) 12
43.6M/s
(CPU) 10
43.4M/s
(WebGL) 10
43.4M/s
(WebGL) 14
43.3M/s
(WebGL) 16
43.1M/s
(WebGL) 17
42.8M/s
(CPU) 18
42.4M/s
(CPU) 13
41.5M/s
(CPU) 11
41.5M/s
(CPU) 9
41.4M/s
(WebGL) 18
41.3M/s
(WebGL) 9
40.8M/s
(CPU) 15
40.3M/s
(CPU) 17
40.1M/s
(CPU) 16
39.9M/s
(WebGL) 11
39.2M/s
(WebGL) 15
37.2M/s
(WebGL) 12
36.5M/s
Async wait for init (WebGL)
2.7M/s
Async wait for init (CPU)
2.4M/s
Async wait for init (WASM)
1.8M/s
(CPU) AddGatherReshape_channelMultiplier_1
463K/s
(WebGL) AddGatherReshape_channelMultiplier_1
325K/s
(WASM) AddGatherReshape_channelMultiplier_1
25K/s
(WebGL) AddGatherReshape_channelMultiplier_2
2K/s
(WebGL) AddGatherReshape_channelMultiplier_4
2K/s
(WebGL) AddGatherReshape_channelMultiplier_8
1K/s
(WebGL) SplitReshapeGatherConcat_channelMultiplier_2
423/s
(WebGL) SplitReshapeGatherConcat_channelMultiplier_1
384/s
(WebGL) SplitReshapeGatherConcat_channelMultiplier_4
320/s
(WebGL) SplitReshapeGatherConcat_channelMultiplier_8
259/s
(CPU) AddGatherReshape_channelMultiplier_2
23/s
(WASM) AddGatherReshape_channelMultiplier_2
22/s
(WASM) SplitReshapeGatherConcat_channelMultiplier_1
22/s
(WASM) SplitReshapeGatherConcat_channelMultiplier_2
20/s
(WASM) AddGatherReshape_channelMultiplier_4
16/s
(CPU) AddGatherReshape_channelMultiplier_4
16/s
(WASM) SplitReshapeGatherConcat_channelMultiplier_4
15/s
(CPU) SplitReshapeGatherConcat_channelMultiplier_1
11/s
(WASM) AddGatherReshape_channelMultiplier_8
10/s
(WASM) SplitReshapeGatherConcat_channelMultiplier_8
10/s
(CPU) SplitReshapeGatherConcat_channelMultiplier_2
10/s
(CPU) AddGatherReshape_channelMultiplier_8
9/s
(CPU) SplitReshapeGatherConcat_channelMultiplier_4
8/s
(CPU) SplitReshapeGatherConcat_channelMultiplier_8
6/s
View exact numbers
Test name Executions per second
(CPU) 14 44,277,356 Ops/sec
(WebGL) 13 44,093,060 Ops/sec
(CPU) 12 43,552,816 Ops/sec
(CPU) 10 43,437,492 Ops/sec
(WebGL) 10 43,437,132 Ops/sec
(WebGL) 14 43,261,352 Ops/sec
(WebGL) 16 43,067,828 Ops/sec
(WebGL) 17 42,785,684 Ops/sec
(CPU) 18 42,360,580 Ops/sec
(CPU) 13 41,508,076 Ops/sec
(CPU) 11 41,506,492 Ops/sec
(CPU) 9 41,417,052 Ops/sec
(WebGL) 18 41,294,776 Ops/sec
(WebGL) 9 40,784,352 Ops/sec
(CPU) 15 40,282,572 Ops/sec
(CPU) 17 40,063,412 Ops/sec
(CPU) 16 39,861,776 Ops/sec
(WebGL) 11 39,156,728 Ops/sec
(WebGL) 15 37,186,940 Ops/sec
(WebGL) 12 36,532,016 Ops/sec
Async wait for init (WebGL) 2,737,472 Ops/sec
Async wait for init (CPU) 2,353,033 Ops/sec
Async wait for init (WASM) 1,760,942 Ops/sec
(CPU) AddGatherReshape_channelMultiplier_1 462,723 Ops/sec
(WebGL) AddGatherReshape_channelMultiplier_1 325,224 Ops/sec
(WASM) AddGatherReshape_channelMultiplier_1 24,690 Ops/sec
(WebGL) AddGatherReshape_channelMultiplier_2 1,798 Ops/sec
(WebGL) AddGatherReshape_channelMultiplier_4 1,781 Ops/sec
(WebGL) AddGatherReshape_channelMultiplier_8 1,279 Ops/sec
(WebGL) SplitReshapeGatherConcat_channelMultiplier_2 423 Ops/sec
(WebGL) SplitReshapeGatherConcat_channelMultiplier_1 384 Ops/sec
(WebGL) SplitReshapeGatherConcat_channelMultiplier_4 320 Ops/sec
(WebGL) SplitReshapeGatherConcat_channelMultiplier_8 259 Ops/sec
(CPU) AddGatherReshape_channelMultiplier_2 23 Ops/sec
(WASM) AddGatherReshape_channelMultiplier_2 22 Ops/sec
(WASM) SplitReshapeGatherConcat_channelMultiplier_1 22 Ops/sec
(WASM) SplitReshapeGatherConcat_channelMultiplier_2 20 Ops/sec
(WASM) AddGatherReshape_channelMultiplier_4 16 Ops/sec
(CPU) AddGatherReshape_channelMultiplier_4 16 Ops/sec
(WASM) SplitReshapeGatherConcat_channelMultiplier_4 15 Ops/sec
(CPU) SplitReshapeGatherConcat_channelMultiplier_1 11 Ops/sec
(WASM) AddGatherReshape_channelMultiplier_8 10 Ops/sec
(WASM) SplitReshapeGatherConcat_channelMultiplier_8 10 Ops/sec
(CPU) SplitReshapeGatherConcat_channelMultiplier_2 10 Ops/sec
(CPU) AddGatherReshape_channelMultiplier_8 9 Ops/sec
(CPU) SplitReshapeGatherConcat_channelMultiplier_4 8 Ops/sec
(CPU) SplitReshapeGatherConcat_channelMultiplier_8 6 Ops/sec