interpret caliper results

for caliper results like

20% Scenario{vm=java, trial=1, benchmark=WithGettersAndSetters, memoryMax=-Xmx512M} 36.39 ns; ?=0.26 ns @ 3 trials, allocated 1 instances for a total of 32B
36.39 ns is median of time it took in each reps
?=0.26 is std dev

and use reps parameter passed to timeMethod to iterate over computation to benchmark.

Advertisements

crawling

https://mailman.cs.umd.edu/mailman/private/javamemorymodel-discussion/2011-June/000518.html
    |
    |
    V
http://www.infoq.com/articles/memory_barriers_jvm_concurrency


http://www.google.com/search?&q=memory+barriers
    |
    |
    V
http://www.rdrop.com/users/paulmck/scalability/paper/whymb.2009.04.05a.pdf
    |
    |
    V
http://www.rdrop.com/users/paulmck/
    |
    |
    V
http://paulmck.livejournal.com/
    |
    |
    V
http://kernel.org/pub/linux/kernel/people/paulmck/perfbook/perfbook.html