Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Real data on cache working set

Author: Tom Kerrigan

Date: 12:02:46 08/24/04

Go up one level in this thread


On August 24, 2004 at 10:04:28, Robert Hyatt wrote:

>>Are you saying that Anandtech's measurements are not real data?
>
>I am saying I have no idea what they are reporting.  I haven't read it and don't
>really care...

Oh, that's grand. You say things like:

"I hate to cloud all the disinformation here with real data"

and

"It took a good bit of time to get a company to help me run the test.  I didn't
do it 'lightly'.  Unlike most of your tests..."

You call Anandtech's and my data disinformation, and that we ran the tests
'lightly' (whatever that means), and in the same breath you say that you can't
be bothered to even look at or care about the data.

That's terrific, you have no moral qualms about badmouthing data that you
haven't even looked at. That's real first-class behavior.

As for your wonderful "real data," I thought of an explanation for it that also
agrees with my data, namely, most of the cache misses are warming the cache for
each new position. If you calculate the cache misses per node after the cache is
already warmed up (using the method I suggested) you'd have an idea of how much
memory Crafty when it's in the middle of a search.

-Tom



This page took 0 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.