From: tmoran@acm.org
Subject: Re: multicore-multithreading benchmarks
Date: Thu, 21 Dec 2006 20:48:14 -0600
Date: 2006-12-21T20:48:14-06:00 [thread overview]
Message-ID: <g8OdnXUIoJrz2hbYnZ2dnUVZ_v63nZ2d@comcast.com> (raw)
Karl Nyberg was kind enough to run the N-CPU quicksort test on his Sun
"Try and Buy" evaluation T1000, with 8 cores, 4 threads per core and got
the results below. For the large-N cases there is a significant speedup
when N becomes a larger power of two. That's logical since the number of
partitions at any given time is a power of two (approximately, since the
split may not have been exactly even).
17:43 1646 x: ./trysortn
CPUs N= 1000 N= 10000 N= 100000 N= 1000000
1 0.001456250 0.018297000 0.233244500 2.654417750
2 0.001455750 0.010290500 0.129741000 1.399496750
3 0.001457000 0.009199000 0.125779500 1.397285750
4 0.001456750 0.007722000 0.126326000 0.787722000
5 0.001458250 0.007488000 0.099067250 0.780770500
6 0.001456500 0.007534000 0.094341500 0.754744250
7 0.001457500 0.006942000 0.094361000 0.744371500
8 0.001456750 0.007131000 0.083258000 0.693531250
17:43 1647 x: ./trysortn
CPUs N= 1000 N= 10000 N= 100000 N= 1000000
1 0.001419750 0.018249250 0.227546000 2.614201000
2 0.001420750 0.010317750 0.120750000 1.363506500
3 0.001418000 0.009661250 0.103181250 1.078301000
4 0.001421500 0.006736750 0.097727750 0.760205250
5 0.001418250 0.006691750 0.092696000 0.728165250
6 0.001418250 0.006635500 0.076543750 0.663240250
7 0.001418250 0.006858750 0.077237750 0.598086250
8 0.001422000 0.007047500 0.074811750 0.481415250
17:44 1648 x: ./trysortn
CPUs N= 1000 N= 10000 N= 100000 N= 1000000
1 0.001445750 0.018350750 0.233402000 2.719813750
2 0.001444250 0.010359250 0.126730750 1.437119750
3 0.001445250 0.008780250 0.125607250 1.437211500
4 0.001443500 0.006597250 0.125789000 1.150943750
5 0.001445500 0.006642000 0.102507250 1.132046500
6 0.001444000 0.006687000 0.100985000 0.937661500
7 0.001445250 0.007033000 0.096396000 0.929958500
8 0.001444000 0.007120500 0.091150250 0.870250250
next reply other threads:[~2006-12-22 2:48 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-12-22 2:48 tmoran [this message]
2006-12-22 6:14 ` multicore-multithreading benchmarks tmoran
2006-12-22 17:12 ` Colin Paul Gloster
2006-12-23 6:50 ` tmoran
2006-12-26 21:58 ` Chip Orange
2006-12-27 8:57 ` tmoran
2006-12-27 17:53 ` Chip Orange
2006-12-27 18:46 ` Jeffrey Creem
2006-12-27 19:51 ` tmoran
replies disabled
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox