Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Sempron vs. Athlon 64: Proof that Crafty's working set is < 256k

Author: Robert Hyatt

Date: 20:43:32 08/22/04

Go up one level in this thread


On August 22, 2004 at 23:37:43, Tom Kerrigan wrote:

>On August 22, 2004 at 23:00:32, Robert Hyatt wrote:
>
>>On August 22, 2004 at 22:48:47, Tom Kerrigan wrote:
>>
>>>On August 22, 2004 at 22:44:16, Robert Hyatt wrote:
>>>
>>>>Aha.  OK.  I just ran a test with (a) no egtables, (b) no compiled in threads
>>>>and I get a running size of about 5.75mb, with 3.75mb for the hash/hashp tables,
>>>
>>>>Who looks like an idiot???
>>>
>>>You.
>>
>>Me?  For running a full-blown crafty version?
>>
>>That's what I thought.
>
>No, you're not an idiot for that. (Duh.) You're an idiot for not knowing if your
>program takes ~1MB of memory or ~16MB of memory. I mean, seriously, you're off
>by more than an order of magnitude here.

No I'm not.  I simply _always_ run a full version, supporting threads, endgame
tables, everything.  Never occurs to me to test with something that I don't
actually use in real games...



>
>You're also an idiot for giving me a hard time about this:
>
>"Again, do you believe that the default code, with 4 megs of hash/phash,
>_really_ runs in 5M of RAM?  I don't.  I did run it under linux and got
>something that seems more reasonable, namely 20M. ... Who looks like an idiot???
>... The one who really _knows_ the program or the one who makes wild guesses
>about the program???"
>
>Yeah, you really KNOW your program when you FORGET that 3/4 of the memory you
>use is devoted to endgame database caches and thread structures. And I'm really
>making "wild guesses" about your program by looking at numbers in the Windows
>Task Manager.


the "thread data" is definitely a part of the program's working set.  It uses it
like crazy.  I mentioned it more than once, never thinking anyone was running on
a version without it.  As I never do that until I thought about trying it to see
if it would get closer to your numbers.

BTW 3/4 is wrong.  On my box I use 300 megs for the endgame decompression
indices.  I was _hardly_ counting that in my calculations.

Now, unless something further comes up, I'll leave the last word to you.  I
don't see much point in continuing to argue about working set size without some
_real_ data.


>
>-Tom



This page took 0.01 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.