[RC5] Fwd: Re: choose.dat
emperor at slic.com
Tue Mar 3 20:51:46 EST 1998
==================BEGIN FORWARDED MESSAGE==================
>Received: from Markgarry at aol.com
> by imo15.mx.aol.com (IMOv13.ems) id VHKTa03502;
> Tue, 3 Mar 1998 10:28:21 -0500 (EST)
>From: Markgarry <Markgarry at aol.com>
>Message-ID: <1d32c3b5.34fc2197 at aol.com>
>Date: Tue, 3 Mar 1998 10:28:21 EST
> I've been getting conflicting info on the GARSP program setup from
> various people.
> Simple question: Will the 90mb Choose.dat (23) a) run on a machine
> with 64mb of RAM, and b) is it really that much faster than the 11mb (20)
> .dat file?
b) it depends if you use it all. When running GARSP, you either specify
MAXBITS, or just accept the default of MAXBITS=20. If you are specifying
then you are going to use about 93 megs for the choose array. This means
you'll use virtual RAM. In some cases this is better than setting
because you will in fact not need to swap too often. But just as often you
may find the ~7% speed gain for rasiing MAXBITS from 22 to 23 is offset by
speed lost due to swapping the extra portion of the choose array.
I would RECOMMEND using MAXBITS=22 since this would minimize any swapping
what is needed for multitasking (i.e., none if GARSP is running by itself).
The MAXBITS=22 setting can be used with the choose_23.dat file, but it only
reads in half of the data, and thus uses 47 megs.
Does this help?
===================END FORWARDED MESSAGE===================
Roy Wilson <emperor at slic.com> <ICQ 8094267>
Lat: 44.850959 Lon: -74.40286 [+/- 6']
To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net
rc5-digest subscribers replace rc5 with rc5-digest
More information about the rc5