ECJ-INTEREST-L Archives

February 2006

ECJ-INTEREST-L@LISTSERV.GMU.EDU

Options: Use Monospaced Font
Show HTML Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Chris Ellingwood <[log in to unmask]>
Reply To:
ECJ Evolutionary Computation Toolkit <[log in to unmask]>
Date:
Mon, 6 Feb 2006 19:43:25 -0500
Content-Type:
text/plain
Parts/Attachments:
text/plain (70 lines)
Thanks for the tip. I gave it a try, but didn't note a significant 
performance difference. 


-XX:+AggressiveHeap -Xmx1400m
Run Time:527.39 -> 100.00%
Setup:54.20 -> 10.28%
Evaluate:144.12 -> 27.33%
PostEval Stats:23.50 -> 4.46%
Optimization:0.00 -> 0.00%
Breed:303.36 -> 57.52%
CheckPoint Stats:0.00 -> 0.00%

-XX:+AggressiveHeap -Xmx1400m -Xms1400m 
Run Time:529.69 -> 100.00%
Setup:57.86 -> 10.92%
Evaluate:143.68 -> 27.13%
PostEval Stats:23.64 -> 4.46%
Optimization:0.00 -> 0.00%
Breed:302.32 -> 57.07%
CheckPoint Stats:0.00 -> 0.00%

----------

On Thu, 2 Feb 2006 10:31:19 +1100, Shane Magrath 
<[log in to unmask]> wrote:

>Why not run the VM with a fixed amount of heap memory?
>
>EG: -Xms1500m -Xmx1500m [-server] # fix the heap size at 1.5GB
>
>Advantages are:
>(a) Dynamic heap resizing is avoided thereby
>avoiding issues to do with object copies, heap fragmentation, 
generational gc etc
>(b) Garbage collection is avoided as long as possible.
>
>I like to run my VM heap fixed at about 75-85% of primary memory.
>
>I would have thought the IO costs to do with logging, perhaps check-
pointing, etc
>would have been significant, perhaps even dominant. Discounting, of 
course fitness evaluation for
>"non-trivial" problems.
>
>>
>
>--
>Shane Magrath
>Doctoral Researcher
>University of Technology Sydney
>PH: 61 2 9514 2527
>
>
>
>--
>UTS CRICOS Provider Code:  00099F
>DISCLAIMER: This email message and any accompanying attachments may 
contain
>confidential information.  If you are not the intended recipient, do not
>read, use, disseminate, distribute or copy this message or attachments.  
If
>you have received this message in error, please notify the sender
>immediately and delete this message. Any views expressed in this message
>are those of the individual sender, except where the sender expressly, and
>with authority, states them to be the views the University of Technology,
>Sydney. Before opening any attachments, please check them for viruses and
>defects.
>

ATOM RSS1 RSS2