Release The Kraken

Here are my results (Win7 64b, Core2 Duo 2.4 GHz):<p><pre><code> Minefield 4 b6 9737.7ms Opera 10.62 14490.8ms Chrome 7.0.517.5 dev 18375.4ms Firefox 4 Beta 5 21721.1ms Safari 5.0 22168.6ms Firefox 3.6.9 30053.3ms Explorer 9 PP4 64817.2ms </code></pre> Though, usual caveat, it's kind of expected that the creator of the benchmark will perform the best (see my older comments on browser benchmarks [1][2]).<p>What's probably more interesting is the order of other browsers (for which this benchmark should be less tuned) - again Opera performed very well (and better than Chrome).<p>[1] <a href="http://news.ycombinator.com/item?id=1676456" rel="nofollow">http://news.ycombinator.com/item?id=1676456</a><p>[2] <a href="http://news.ycombinator.com/item?id=1458529" rel="nofollow">http://news.ycombinator.com/item?id=1458529</a>

I see what's going on here. Mozilla's trace compilation approach has a lot of startup overhead but produces extremely fast optimized code for inner loops. Therefore, their benchmark includes compute-intensive tests with long-running inner loops to amortize out the costs of JIT compilation and focus on the raw speed of the generated code, where they beat Chrome.<p>(Note that I'm not saying this is unfair or biased at all. This is a good set of benchmarks to have.)

I think JS performance on Firefox has now reached the point where its good enough. All I ask for now is parts of the UI to become multithreaded so as to prevent the random issues one gets when having in the neighborhood of 150 tabs.

Firefox4 is faster than chrome here on ubuntu lucid x86_64 on an i5 with 8g of RAM.<p>Results for firefox 4 beta7:<p><a href="http://krakenbenchmark.mozilla.com/kraken-1.0/results.html?%7B%22v%22:%20%22kraken-1.0%22,%20%22ai-astar%22:%5B822,823,833,918,808,806,930,893,794,906%5D,%22audio-beat-detection%22:%5B950,945,939,964,959,994,1015,855,929,985%5D,%22audio-dft%22:%5B524,522,533,546,576,557,574,507,532,500%5D,%22audio-fft%22:%5B896,808,862,908,868,886,803,893,904,887%5D,%22audio-oscillator%22:%5B400,405,398,396,431,420,416,408,396,426%5D,%22imaging-gaussian-blur%22:%5B669,711,679,675,672,675,686,687,761,679%5D,%22imaging-darkroom%22:%5B287,261,267,265,267,262,261,265,260,263%5D,%22imaging-desaturate%22:%5B631,617,666,628,615,621,637,659,626,648%5D,%22json-parse-financial%22:%5B270,269,260,258,262,263,258,279,259,275%5D,%22json-stringify-tinderbox%22:%5B76,77,81,73,75,77,71,71,70,74%5D,%22stanford-crypto-aes%22:%5B455,481,446,439,457,471,430,431,439,455%5D,%22stanford-crypto-ccm%22:%5B347,376,358,341,363,391,334,351,361,345%5D,%22stanford-crypto-pbkdf2%22:%5B876,854,743,725,858,920,811,754,823,834%5D,%22stanford-crypto-sha256-iterative%22:%5B125,122,119,126,119,133,121,118,119,119%5D%7D" rel="nofollow">http://krakenbenchmark.mozilla.com/kraken-1.0/results.html?%...</a><p>Results for chrome 7.0.517.0 dev:<p><a href="http://krakenbenchmark.mozilla.com/kraken-1.0/results.html?%7B%22v%22:%20%22kraken-1.0%22,%20%22ai-astar%22:%5B730,728,722,706,782,718,707,712,709,693%5D,%22audio-beat-detection%22:%5B1316,1243,1286,1311,1362,1245,1310,1309,1286,1307%5D,%22audio-dft%22:%5B1954,1820,1933,2047,2151,1838,2022,1901,1986,1868%5D,%22audio-fft%22:%5B1191,1156,1314,1199,1281,1148,1181,1174,1243,1199%5D,%22audio-oscillator%22:%5B470,436,442,436,461,425,472,442,442,455%5D,%22imaging-gaussian-blur%22:%5B3258,2979,3169,2845,3094,2934,3424,2943,2953,3057%5D,%22imaging-darkroom%22:%5B1356,1340,1375,1353,1359,1391,1417,1384,1369,1387%5D,%22imaging-desaturate%22:%5B1458,1383,1429,1472,1366,1387,1417,1420,1364,1354%5D,%22json-parse-financial%22:%5B241,245,252,250,276,241,272,249,236,232%5D,%22json-stringify-tinderbox%22:%5B177,203,194,221,177,180,176,176,180,175%5D,%22stanford-crypto-aes%22:%5B110,120,108,160,111,117,108,112,112,133%5D,%22stanford-crypto-ccm%22:%5B102,101,92,114,98,92,104,95,97,114%5D,%22stanford-crypto-pbkdf2%22:%5B286,224,244,264,231,232,255,241,225,243%5D,%22stanford-crypto-sha256-iterative%22:%5B100,99,102,145,110,96,100,102,96,98%5D%7D" rel="nofollow">http://krakenbenchmark.mozilla.com/kraken-1.0/results.html?%...</a>

The beat detection section was most interesting to me, as I spent a lot of time trying (and mostly failing) to do reliable beat detection when I was in high school for an XMMS plugin. It amazes me that this is now possible in a web browser.

It'd be nice if one of the JS benchmarks actually POST'd the results back to the server, with current browser, and had a UI that allowed comparing the results for different browsers on different tests.

On mine, I notice Opera performs the worst by more than double.<p>(Win7, AMD Athlon XP2400 2.0GHz)<p><pre><code> Chrome 6.0.472.55 beta: 46946.5ms Opera 10.62: 234444.6ms IE9: 71864.8ms</code></pre>

Is there any documentation? What are the components of this benchmark?