[Webkit-unassigned] [Bug 70995] The GC should be parallel
bugzilla-daemon at webkit.org
bugzilla-daemon at webkit.org
Sun Oct 30 01:18:11 PDT 2011
https://bugs.webkit.org/show_bug.cgi?id=70995
--- Comment #23 from Filip Pizlo <fpizlo at apple.com> 2011-10-30 01:18:11 PST ---
Performance retested after merging/fixing/etc.
Benchmark report for SunSpider, V8, and Kraken.
VMs tested:
"TipOfTree" at /Volumes/Data/pizlo/quinary/OpenSource/WebKitBuild/Release/jsc
"ParallelGC" at /Volumes/Data/pizlo/senary/OpenSource/WebKitBuild/Release/jsc
Collected 12 samples per benchmark/VM, with 4 VM invocations per benchmark. Emitted a call to gc() between sample
measurements. Used 1 benchmark iteration per VM invocation for warm-up. Used the jsc-specific preciseTime()
function to get microsecond-level timing. Reporting benchmark execution times with 95% confidence intervals in
milliseconds.
TipOfTree ParallelGC
SunSpider:
3d-cube 7.6899+-0.2978 7.5047+-0.1976 might be 1.0247x faster
3d-morph 7.9344+-0.2444 7.6294+-0.1314 might be 1.0400x faster
3d-raytrace 7.7543+-0.3009 7.4545+-0.1448 might be 1.0402x faster
access-binary-trees 1.6432+-0.0460 1.5978+-0.0653 might be 1.0284x faster
access-fannkuch 6.5333+-0.0883 ? 6.5941+-0.1215 ?
access-nbody 3.8733+-0.1735 3.6594+-0.0649 might be 1.0585x faster
access-nsieve 2.6750+-0.0760 2.5717+-0.0545 might be 1.0402x faster
bitops-3bit-bits-in-byte 1.2764+-0.0264 ! 1.3604+-0.0318 ! definitely 1.0657x slower
bitops-bits-in-byte 2.3180+-0.0668 ? 2.3830+-0.0528 ? might be 1.0280x slower
bitops-bitwise-and 3.4243+-0.0911 3.3971+-0.1047
bitops-nsieve-bits 5.3360+-0.0961 ? 5.4479+-0.1615 ? might be 1.0210x slower
controlflow-recursive 2.1010+-0.0347 ? 2.1506+-0.0343 ? might be 1.0236x slower
crypto-aes 7.4222+-0.2491 ? 7.6064+-0.1827 ? might be 1.0248x slower
crypto-md5 2.7823+-0.0877 2.7117+-0.0968 might be 1.0260x faster
crypto-sha1 2.4500+-0.0612 ? 2.4609+-0.0447 ?
date-format-tofte 10.1750+-0.3408 ? 10.3077+-0.3219 ? might be 1.0130x slower
date-format-xparb 9.0613+-0.2235 ! 9.8251+-0.3201 ! definitely 1.0843x slower
math-cordic 6.4947+-0.1313 6.4855+-0.1292
math-partial-sums 7.5811+-0.2072 7.5517+-0.2192
math-spectral-norm 2.6240+-0.0661 ? 2.6337+-0.0613 ?
regexp-dna 11.6862+-0.2992 11.6826+-0.3405
string-base64 4.1872+-0.0378 ? 4.4519+-0.2612 ? might be 1.0632x slower
string-fasta 6.6351+-0.2360 6.5851+-0.2751
string-tagcloud 11.7555+-0.4311 ? 12.0805+-0.3561 ? might be 1.0276x slower
string-unpack-code 20.5635+-0.4911 ? 21.3864+-0.8229 ? might be 1.0400x slower
string-validate-input 5.3534+-0.0821 ^ 5.1113+-0.0723 ^ definitely 1.0474x faster
<arithmetic> * 6.2050+-0.0507 ? 6.2550+-0.0435 ?
<geometric> 4.9988+-0.0332 ? 5.0137+-0.0282 ?
<harmonic> 3.9802+-0.0270 ? 3.9967+-0.0305 ?
TipOfTree ParallelGC
V8:
crypto 73.1870+-0.3514 ? 73.4561+-0.4061 ?
deltablue 178.5365+-0.8240 ^ 174.2394+-0.5655 ^ definitely 1.0247x faster
earley-boyer 91.9127+-0.3644 91.5211+-0.4322
raytrace 62.8774+-0.3963 ? 63.4041+-0.6165 ?
regexp 106.1395+-0.6146 ? 107.0600+-0.6550 ?
richards 126.5740+-0.4962 126.5667+-0.2941
splay 93.7210+-0.4231 ^ 72.9378+-1.0055 ^ definitely 1.2849x faster
<arithmetic> 104.7069+-0.2350 ^ 101.3122+-0.2225 ^ definitely 1.0335x faster
<geometric> * 99.2797+-0.2067 ^ 95.6737+-0.2585 ^ definitely 1.0377x faster
<harmonic> 94.5382+-0.1999 ^ 90.8940+-0.2924 ^ definitely 1.0401x faster
TipOfTree ParallelGC
Kraken:
ai-astar 493.6872+-1.0811 ! 498.7911+-2.1791 ! definitely 1.0103x slower
audio-beat-detection 193.5298+-1.4174 ^ 190.9533+-0.8591 ^ definitely 1.0135x faster
audio-dft 267.3048+-1.8658 ? 269.0041+-2.7896 ?
audio-fft 124.6659+-0.8884 ? 124.9870+-1.0028 ?
audio-oscillator 251.0656+-1.6563 ? 251.5320+-1.1005 ?
imaging-darkroom 403.1887+-2.5335 400.9468+-1.4467
imaging-desaturate 225.4055+-0.4488 ? 225.9516+-1.6251 ?
imaging-gaussian-blur 554.4614+-1.3730 ? 555.1623+-2.2363 ?
json-parse-financial 58.4412+-0.8444 ^ 57.0616+-0.3266 ^ definitely 1.0242x faster
json-stringify-tinderbox 69.5005+-0.6899 ? 70.3635+-1.0300 ? might be 1.0124x slower
stanford-crypto-aes 133.2301+-1.1865 132.4162+-1.3045
stanford-crypto-ccm 100.1346+-0.6399 ? 100.1434+-0.6771 ?
stanford-crypto-pbkdf2 196.4266+-3.2542 ^ 191.2882+-1.0304 ^ definitely 1.0269x faster
stanford-crypto-sha256-iterative 71.3867+-0.5641 71.2631+-0.2770
<arithmetic> * 224.4592+-0.3648 224.2760+-0.4353
<geometric> 177.3997+-0.4300 176.8961+-0.3648
<harmonic> 140.3279+-0.6257 139.7143+-0.4019
TipOfTree ParallelGC
All benchmarks:
<arithmetic> 85.8874+-0.1416 ^ 85.3549+-0.1542 ^ definitely 1.0062x faster
<geometric> 22.5895+-0.0949 22.4835+-0.0785
<harmonic> 7.0084+-0.0466 ? 7.0333+-0.0525 ?
TipOfTree ParallelGC
Geomean of preferred means:
<scaled-result> 51.7098+-0.1562 ^ 51.1993+-0.1660 ^ definitely 1.0100x faster
--
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.
More information about the webkit-unassigned
mailing list