email@hidden wrote:
> My own experiences with the server VM on Windows have been rather
>less positive. I tried it with one of my own programs, in the hope that
>it would run faster. Although it's a desktop application, not a server
>one, it seemed like exactly the sort of thing that would be helped by the
>server VM: a lengthy, highly CPU intensive calculation where a few
>core routines get executed millions of times.
> The result? The execution time with the client VM was very
>consistent, while the time with the server VM varied wildly. Sometimes it
>took exactly the same amount of time as the client VM. Sometimes it was
>about 30% faster. And sometimes it was 10x slower. (Yes, you read that
>correctly. A calculation that took 1 minute with the client VM would
>sometimes take over 10 minutes with the server VM.)
> As you can guess, I stuck with the client VM.
The server VM is intended for processes with long lifetimes. It's not just
that a few routines get executed millions of times. The process itself
typically has to run for a long enough period of time that the overhead of
performing the optimizations gets amortized.
Depending on what the program was doing, and how it was tested, I think
it's entirely possible that the server VM produces widely varying execution
times.
-- GG
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Java-dev mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
http://lists.apple.com/mailman/options/java-dev/email@hidden
This email sent to email@hidden
