GC performance benchmarks

[Y. Srinivas Ramakrishna]

Sorry I missed this thread's prefix....

On 12/22/2010 2:27 AM, Pressenna wrote:
> Hi Dalibor,
>
> I tried Dacapo and it crashes the VM. And when it manages to run it is
> about twice as slow :(

Which JVM are we talking about here, and comparing with what ("twice as slow")?

> I fear my test cases do not represent real world application behaviour.

Did you mean the Da Capo tests? I believe they do stress GC's in various ways;
whether real world, it's probably difficult to generalize, I guess....

You might also try specjbb2005 etc. from spec.org, although how realistic that
is, is also difficult (for me at least) to say.

-- ramki