Benchmark: (ColorfulCakeChen) CNN Channel Shuffler (50c1b285ff7c7f8c656c388365b4043528642d6b) 98

(ColorfulCakeChen) CNN Channel Shuffler (50c1b285ff7c7f8c656c388365b4043528642d6b) 98 (version: 98)

Comparing performance of: Async wait for init (WebGL) vs (WebGL) (G8) ConcatReshapeTransposeReshapeSplit vs (WebGL) (G8) ConcatGatherUnsorted vs (WebGL) (G8) SplitConcatSortedShared vs (WebGL) (G8) ConcatPointwiseConv vs (WebGL) (G4) ConcatReshapeTransposeReshapeSplit vs (WebGL) (G4) ConcatGatherUnsorted vs (WebGL) (G4) SplitConcatSortedShared vs (WebGL) (G4) ConcatPointwiseConv vs (WebGL) (G3) ConcatReshapeTransposeReshapeSplit vs (WebGL) (G3) ConcatGatherUnsorted vs (WebGL) (G3) SplitConcatSortedShared vs (WebGL) (G3) ConcatPointwiseConv vs (WebGL) (G2) ConcatReshapeTransposeReshapeSplit vs (WebGL) (G2) ConcatGatherUnsorted vs (WebGL) (G2) SplitConcatSortedShared vs (WebGL) (G2) ConcatPointwiseConv vs (WebGL) (G1) ConcatReshapeTransposeReshapeSplit vs (WebGL) (G1) ConcatGatherUnsorted vs (WebGL) (G1) SplitConcatSortedShared vs (WebGL) (G1) ConcatPointwiseConv vs Async wait for init (CPU) vs (CPU) (G8) ConcatReshapeTransposeReshapeSplit vs (CPU) (G8) ConcatGatherUnsorted vs (CPU) (G8) SplitConcatSortedShared vs (CPU) (G8) ConcatPointwiseConv vs (CPU) (G4) ConcatReshapeTransposeReshapeSplit vs (CPU) (G4) ConcatGatherUnsorted vs (CPU) (G4) SplitConcatSortedShared vs (CPU) (G4) ConcatPointwiseConv vs (CPU) (G3) ConcatReshapeTransposeReshapeSplit vs (CPU) (G3) ConcatGatherUnsorted vs (CPU) (G3) SplitConcatSortedShared vs (CPU) (G3) ConcatPointwiseConv vs (CPU) (G2) ConcatReshapeTransposeReshapeSplit vs (CPU) (G2) ConcatGatherUnsorted vs (CPU) (G2) SplitConcatSortedShared vs (CPU) (G2) ConcatPointwiseConv vs (CPU) (G1) ConcatReshapeTransposeReshapeSplit vs (CPU) (G1) ConcatGatherUnsorted vs (CPU) (G1) SplitConcatSortedShared vs (CPU) (G1) ConcatPointwiseConv

Created: 3 years ago by: Registered User

Jump to the latest result

HTML Preparation code:

<script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs@4.8.0/dist/tf.min.js"></script>
<script src="https://cdn.jsdelivr.net/npm/@tensorflow/tfjs-backend-wasm/dist/tf-backend-wasm.js"></script>

Script Preparation code:

async function init( backendName = "webgl" ) {
  let jsPerf_CNNChannelShuffle = await import(
    "https://cdn.jsdelivr.net/gh/colorfulcakechen/query-submit-canvas@50c1b285ff7c7f8c656c388365b4043528642d6b/CNN/jsPerf/jsPerf_CNNChannelShuffle.js" );

  jsPerf_CNNChannelShuffle.disposeResources();

  await tf.ready();
  await tf.setBackend( backendName );

  jsPerf_CNNChannelShuffle.init();
  jsPerf_CNNChannelShuffle.testResultSame();
  jsPerf_CNNChannelShuffle.testDifferentDisposeStrategy_All();

  return true;
}

Tests:

Async wait for init (WebGL)
if ( globalThis.initPromiseWebGL && globalThis.initPromiseCPU ) // For re-run benchmark. globalThis.initPromiseWebGL = globalThis.initPromiseCPU = null; if ( !globalThis.initPromiseWebGL ) globalThis.initPromiseWebGL = init( "webgl" ); // Just busy waiting.
(WebGL) (G8) ConcatReshapeTransposeReshapeSplit
globalThis.testSet_110x110x24_g8.test_ConcatReshapeTransposeReshapeSplit();
(WebGL) (G8) ConcatGatherUnsorted
globalThis.testSet_110x110x24_g8.test_ConcatGatherUnsorted();
(WebGL) (G8) SplitConcatSortedShared
globalThis.testSet_110x110x24_g8.test_SplitConcatSortedShared();
(WebGL) (G8) ConcatPointwiseConv
globalThis.testSet_110x110x24_g8.test_ConcatPointwiseConv();
(WebGL) (G4) ConcatReshapeTransposeReshapeSplit
globalThis.testSet_110x110x24_g4.test_ConcatReshapeTransposeReshapeSplit();
(WebGL) (G4) ConcatGatherUnsorted
globalThis.testSet_110x110x24_g4.test_ConcatGatherUnsorted();
(WebGL) (G4) SplitConcatSortedShared
globalThis.testSet_110x110x24_g4.test_SplitConcatSortedShared();
(WebGL) (G4) ConcatPointwiseConv
globalThis.testSet_110x110x24_g4.test_ConcatPointwiseConv();
(WebGL) (G3) ConcatReshapeTransposeReshapeSplit
globalThis.testSet_110x110x24_g3.test_ConcatReshapeTransposeReshapeSplit();
(WebGL) (G3) ConcatGatherUnsorted
globalThis.testSet_110x110x24_g3.test_ConcatGatherUnsorted();
(WebGL) (G3) SplitConcatSortedShared
globalThis.testSet_110x110x24_g3.test_SplitConcatSortedShared();
(WebGL) (G3) ConcatPointwiseConv
globalThis.testSet_110x110x24_g3.test_ConcatPointwiseConv();
(WebGL) (G2) ConcatReshapeTransposeReshapeSplit
globalThis.testSet_110x110x24_g2.test_ConcatReshapeTransposeReshapeSplit();
(WebGL) (G2) ConcatGatherUnsorted
globalThis.testSet_110x110x24_g2.test_ConcatGatherUnsorted();
(WebGL) (G2) SplitConcatSortedShared
globalThis.testSet_110x110x24_g2.test_SplitConcatSortedShared();
(WebGL) (G2) ConcatPointwiseConv
globalThis.testSet_110x110x24_g2.test_ConcatPointwiseConv();
(WebGL) (G1) ConcatReshapeTransposeReshapeSplit
globalThis.testSet_110x110x24_g1.test_ConcatReshapeTransposeReshapeSplit();
(WebGL) (G1) ConcatGatherUnsorted
globalThis.testSet_110x110x24_g1.test_ConcatGatherUnsorted();
(WebGL) (G1) SplitConcatSortedShared
globalThis.testSet_110x110x24_g1.test_SplitConcatSortedShared();
(WebGL) (G1) ConcatPointwiseConv
globalThis.testSet_110x110x24_g1.test_ConcatPointwiseConv();
Async wait for init (CPU)
if ( !globalThis.initPromiseCPU ) globalThis.initPromiseCPU = init( "cpu" ); // Just busy waiting.
(CPU) (G8) ConcatReshapeTransposeReshapeSplit
globalThis.testSet_110x110x24_g8.test_ConcatReshapeTransposeReshapeSplit();
(CPU) (G8) ConcatGatherUnsorted
globalThis.testSet_110x110x24_g8.test_ConcatGatherUnsorted();
(CPU) (G8) SplitConcatSortedShared
globalThis.testSet_110x110x24_g8.test_SplitConcatSortedShared();
(CPU) (G8) ConcatPointwiseConv
globalThis.testSet_110x110x24_g8.test_ConcatPointwiseConv();
(CPU) (G4) ConcatReshapeTransposeReshapeSplit
globalThis.testSet_110x110x24_g4.test_ConcatReshapeTransposeReshapeSplit();
(CPU) (G4) ConcatGatherUnsorted
globalThis.testSet_110x110x24_g4.test_ConcatGatherUnsorted();
(CPU) (G4) SplitConcatSortedShared
globalThis.testSet_110x110x24_g4.test_SplitConcatSortedShared();
(CPU) (G4) ConcatPointwiseConv
globalThis.testSet_110x110x24_g4.test_ConcatPointwiseConv();
(CPU) (G3) ConcatReshapeTransposeReshapeSplit
globalThis.testSet_110x110x24_g3.test_ConcatReshapeTransposeReshapeSplit();
(CPU) (G3) ConcatGatherUnsorted
globalThis.testSet_110x110x24_g3.test_ConcatGatherUnsorted();
(CPU) (G3) SplitConcatSortedShared
globalThis.testSet_110x110x24_g3.test_SplitConcatSortedShared();
(CPU) (G3) ConcatPointwiseConv
globalThis.testSet_110x110x24_g3.test_ConcatPointwiseConv();
(CPU) (G2) ConcatReshapeTransposeReshapeSplit
globalThis.testSet_110x110x24_g2.test_ConcatReshapeTransposeReshapeSplit();
(CPU) (G2) ConcatGatherUnsorted
globalThis.testSet_110x110x24_g2.test_ConcatGatherUnsorted();
(CPU) (G2) SplitConcatSortedShared
globalThis.testSet_110x110x24_g2.test_SplitConcatSortedShared();
(CPU) (G2) ConcatPointwiseConv
globalThis.testSet_110x110x24_g2.test_ConcatPointwiseConv();
(CPU) (G1) ConcatReshapeTransposeReshapeSplit
globalThis.testSet_110x110x24_g1.test_ConcatReshapeTransposeReshapeSplit();
(CPU) (G1) ConcatGatherUnsorted
globalThis.testSet_110x110x24_g1.test_ConcatGatherUnsorted();
(CPU) (G1) SplitConcatSortedShared
globalThis.testSet_110x110x24_g1.test_SplitConcatSortedShared();
(CPU) (G1) ConcatPointwiseConv
globalThis.testSet_110x110x24_g1.test_ConcatPointwiseConv();

Rendered benchmark preparation results:

Suite status: <idle, ready to run>

Previous results

Test case name	Result
Async wait for init (WebGL)
(WebGL) (G8) ConcatReshapeTransposeReshapeSplit
(WebGL) (G8) ConcatGatherUnsorted
(WebGL) (G8) SplitConcatSortedShared
(WebGL) (G8) ConcatPointwiseConv
(WebGL) (G4) ConcatReshapeTransposeReshapeSplit
(WebGL) (G4) ConcatGatherUnsorted
(WebGL) (G4) SplitConcatSortedShared
(WebGL) (G4) ConcatPointwiseConv
(WebGL) (G3) ConcatReshapeTransposeReshapeSplit
(WebGL) (G3) ConcatGatherUnsorted
(WebGL) (G3) SplitConcatSortedShared
(WebGL) (G3) ConcatPointwiseConv
(WebGL) (G2) ConcatReshapeTransposeReshapeSplit
(WebGL) (G2) ConcatGatherUnsorted
(WebGL) (G2) SplitConcatSortedShared
(WebGL) (G2) ConcatPointwiseConv
(WebGL) (G1) ConcatReshapeTransposeReshapeSplit
(WebGL) (G1) ConcatGatherUnsorted
(WebGL) (G1) SplitConcatSortedShared
(WebGL) (G1) ConcatPointwiseConv
Async wait for init (CPU)
(CPU) (G8) ConcatReshapeTransposeReshapeSplit
(CPU) (G8) ConcatGatherUnsorted
(CPU) (G8) SplitConcatSortedShared
(CPU) (G8) ConcatPointwiseConv
(CPU) (G4) ConcatReshapeTransposeReshapeSplit
(CPU) (G4) ConcatGatherUnsorted
(CPU) (G4) SplitConcatSortedShared
(CPU) (G4) ConcatPointwiseConv
(CPU) (G3) ConcatReshapeTransposeReshapeSplit
(CPU) (G3) ConcatGatherUnsorted
(CPU) (G3) SplitConcatSortedShared
(CPU) (G3) ConcatPointwiseConv
(CPU) (G2) ConcatReshapeTransposeReshapeSplit
(CPU) (G2) ConcatGatherUnsorted
(CPU) (G2) SplitConcatSortedShared
(CPU) (G2) ConcatPointwiseConv
(CPU) (G1) ConcatReshapeTransposeReshapeSplit
(CPU) (G1) ConcatGatherUnsorted
(CPU) (G1) SplitConcatSortedShared
(CPU) (G1) ConcatPointwiseConv

Fastest: N/A

Slowest: N/A

Latest run results:

Run details: (Test run date: one year ago)

User agent: Mozilla/5.0 (X11; CrOS x86_64 14541.0.0) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/114.0.0.0 Safari/537.36

Browser/OS: Chrome 114 on Chrome OS 14541.0.0

View result in a separate tab

Test name	Executions per second
Async wait for init (WebGL)	600999.6 Ops/sec
(CPU) (G8) ConcatGatherUnsorted	21.2 Ops/sec
(CPU) (G8) SplitConcatSortedShared	3.7 Ops/sec
(CPU) (G8) ConcatPointwiseConv	9.0 Ops/sec
(CPU) (G4) ConcatReshapeTransposeReshapeSplit	3.2 Ops/sec
(CPU) (G4) ConcatGatherUnsorted	20.4 Ops/sec
(CPU) (G4) SplitConcatSortedShared	3.4 Ops/sec
(CPU) (G4) ConcatPointwiseConv	10.1 Ops/sec
(CPU) (G3) ConcatReshapeTransposeReshapeSplit	3.3 Ops/sec
(CPU) (G3) ConcatGatherUnsorted	23.1 Ops/sec
(CPU) (G3) SplitConcatSortedShared	3.2 Ops/sec
(CPU) (G3) ConcatPointwiseConv	11.2 Ops/sec
(CPU) (G2) ConcatReshapeTransposeReshapeSplit	3.4 Ops/sec
(CPU) (G2) ConcatGatherUnsorted	22.2 Ops/sec
(CPU) (G2) SplitConcatSortedShared	3.1 Ops/sec
(CPU) (G2) ConcatPointwiseConv	12.6 Ops/sec
(CPU) (G1) ConcatReshapeTransposeReshapeSplit	30.7 Ops/sec
(CPU) (G1) ConcatGatherUnsorted	25.3 Ops/sec
(CPU) (G8) ConcatReshapeTransposeReshapeSplit	3.0 Ops/sec
Async wait for init (CPU)	3589551.5 Ops/sec
(WebGL) (G1) ConcatPointwiseConv	1347.9 Ops/sec
(WebGL) (G1) SplitConcatSortedShared	62.0 Ops/sec
(WebGL) (G8) ConcatReshapeTransposeReshapeSplit	240.7 Ops/sec
(WebGL) (G8) ConcatGatherUnsorted	78.1 Ops/sec
(WebGL) (G8) SplitConcatSortedShared	214.5 Ops/sec
(WebGL) (G8) ConcatPointwiseConv	431.7 Ops/sec
(WebGL) (G4) ConcatReshapeTransposeReshapeSplit	591.2 Ops/sec
(WebGL) (G4) ConcatGatherUnsorted	190.6 Ops/sec
(WebGL) (G4) SplitConcatSortedShared	145.0 Ops/sec
(WebGL) (G4) ConcatPointwiseConv	414.5 Ops/sec
(CPU) (G1) SplitConcatSortedShared	3.3 Ops/sec
(WebGL) (G3) ConcatReshapeTransposeReshapeSplit	421.2 Ops/sec
(WebGL) (G3) SplitConcatSortedShared	178.5 Ops/sec
(WebGL) (G3) ConcatPointwiseConv	763.4 Ops/sec
(WebGL) (G2) ConcatReshapeTransposeReshapeSplit	464.1 Ops/sec
(WebGL) (G2) ConcatGatherUnsorted	325.6 Ops/sec
(WebGL) (G2) SplitConcatSortedShared	269.8 Ops/sec
(WebGL) (G2) ConcatPointwiseConv	913.4 Ops/sec
(WebGL) (G1) ConcatReshapeTransposeReshapeSplit	771.1 Ops/sec
(WebGL) (G1) ConcatGatherUnsorted	2759.0 Ops/sec
(WebGL) (G3) ConcatGatherUnsorted	313.8 Ops/sec
(CPU) (G1) ConcatPointwiseConv	14.5 Ops/sec

Autogenerated LLM Summary (model llama3.2:3b, generated one month ago):

LLMs can make mistakes. Check important info.

Related benchmarks:

(ColorfulCakeChen) CNN Channel Shuffler (50c1b285ff7c7f8c656c388365b4043528642d6b) 98 (version: 98)

Created: 3 years ago by: Registered User

Jump to the latest result

Async wait for init (WebGL)

(WebGL) (G8) ConcatReshapeTransposeReshapeSplit

(WebGL) (G8) ConcatGatherUnsorted

(WebGL) (G8) SplitConcatSortedShared

(WebGL) (G8) ConcatPointwiseConv

(WebGL) (G4) ConcatReshapeTransposeReshapeSplit

(WebGL) (G4) ConcatGatherUnsorted

(WebGL) (G4) SplitConcatSortedShared

(WebGL) (G4) ConcatPointwiseConv

(WebGL) (G3) ConcatReshapeTransposeReshapeSplit

(WebGL) (G3) ConcatGatherUnsorted

(WebGL) (G3) SplitConcatSortedShared

(WebGL) (G3) ConcatPointwiseConv

(WebGL) (G2) ConcatReshapeTransposeReshapeSplit

(WebGL) (G2) ConcatGatherUnsorted

(WebGL) (G2) SplitConcatSortedShared

(WebGL) (G2) ConcatPointwiseConv

(WebGL) (G1) ConcatReshapeTransposeReshapeSplit

(WebGL) (G1) ConcatGatherUnsorted

(WebGL) (G1) SplitConcatSortedShared

(WebGL) (G1) ConcatPointwiseConv

Async wait for init (CPU)

(CPU) (G8) ConcatReshapeTransposeReshapeSplit

(CPU) (G8) ConcatGatherUnsorted

(CPU) (G8) SplitConcatSortedShared

(CPU) (G8) ConcatPointwiseConv

(CPU) (G4) ConcatReshapeTransposeReshapeSplit

(CPU) (G4) ConcatGatherUnsorted

(CPU) (G4) SplitConcatSortedShared

(CPU) (G4) ConcatPointwiseConv

(CPU) (G3) ConcatReshapeTransposeReshapeSplit

(CPU) (G3) ConcatGatherUnsorted

(CPU) (G3) SplitConcatSortedShared

(CPU) (G3) ConcatPointwiseConv

(CPU) (G2) ConcatReshapeTransposeReshapeSplit

(CPU) (G2) ConcatGatherUnsorted

(CPU) (G2) SplitConcatSortedShared

(CPU) (G2) ConcatPointwiseConv

(CPU) (G1) ConcatReshapeTransposeReshapeSplit

(CPU) (G1) ConcatGatherUnsorted

(CPU) (G1) SplitConcatSortedShared

(CPU) (G1) ConcatPointwiseConv

Suite status: <idle, ready to run>

Experimental features:

Fastest: N/A

Slowest: N/A

Autogenerated LLM Summary (model llama3.2:3b, generated one month ago):