on master help sign in
Update benchmark results
a7df4baBryan O'Sullivan
aeson.cabal+14−5
@@ -18,15 +18,24 @@ description:
18 and high performance.
18 and high performance.
19 .
19 .
20 Parsing performance with GHC 6.12.3 on a late 2010 MacBook Pro
20 Parsing performance with GHC 6.12.3 on a late 2010 MacBook Pro
21− (2.66GHz Core i7), for different message sizes:
21+ (2.66GHz Core i7), for mostly-English tweets from Twitter's JSON
22+ search API:
22 .
23 .
23− * 854 bytes: 15308 msg\/sec (12.5 MB/sec)
24+ * 854 bytes: 21054 msg\/sec (17.1 MB/sec)
24 .
25 .
25− * 6.4 KB: 2807 msg\/sec (17.6 MB/sec)
26+ * 6.4 KB: 4545 msg\/sec (28.6 MB/sec)
26 .
27 .
27− * 31.2 KB: 493 msg\/sec (15.0 MB/sec)
28+ * 31.2 KB: 856 msg\/sec (26.1 MB/sec)
28 .
29 .
29− * 61.5 KB: 236 msg\/sec (14.2 MB/sec)
30+ * 61.5 KB: 403 msg\/sec (24.2 MB/sec)
31+ .
32+ Handling heavily-escaped text is a little more work. Here is
33+ parsing performance with Japanese tweets, where much of the text
34+ is entirely Unicode-escaped:
35+ .
36+ * 14.6 KB: 1250 msg\/sec (17.9 MB/sec)
37+ .
38+ * 44.1 KB: 363 msg\/sec (15.6 MB/sec)
30 .
39 .
31 Encoding performance on the same machine and data:
40 Encoding performance on the same machine and data:
32 .
41 .