Date:         Thu, 5 Jun 2003 13:20:03 -0700
Reply-To:     philrack@MINEQUEST.COM
Sender:       "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From:         Phil Rack <philrack@MINEQUEST.COM>
Subject:      Re: Win2K and Dual Processors
Comments: To: "William W. Viergever" <wwvierg@attglobal.net>
In-Reply-To:  <5.2.1.1.2.20030605130152.00af63c0@pop3.attglobal.net>
Content-Type: text/plain; charset=ISO-8859-1

Excellent News!  I'm getting ready to purchase a dual Athlon MP setup (still
doing research and cost planning) and wanted to get the most bang for the buck
using SAS.

Thanks,

Phil


Quoting "William W. Viergever" <wwvierg@attglobal.net>:

> At 12:18 PM 06/05/2003 -0700, Phil Rack wrote:
>
> >SAS-L'rs
> >
> >I was wondering... if I was using a Win2K Workstation with dual processors
> and
> >had two local SAS sessions running, would SAS (or the OS) be smart enough
> to
> >use both processors for each SAS job?
> >
> >TIA
> >
> >Phil
> >
> >Phil Rack
> >SAS Consulting and Contracting Services
> >www.MineQuest.Com
> >Tel: (614) 457-3714
> >Fax: (614) 457-9394
> >E-mail: philrack@minequest.com
>
>
>
> Phil:
>
> I have a dualie Athlon, running v8.2 under W2K.
>
> I just started two jobs, running against two different annual Medicaid
> inpatient claims datasets (both jobs run from same directory, and both
> datasets live, albeit on another drive, in the same directory) and opened
> up Task Manager and switched to the Performance tab.
>
> Both CPUs, visually, appear to be being used evenly: both at near 19-22%
> (freqs) at the beginning and are now chugging along at about 4-8% usage
> (summaries?).
>
>
> The jobs are Freqs and Summaries:
>
> proc freq data = &dsn1..&file1 ;
>     where cross = 'N' ;
>     tables plancode*clmtype*procind*accmcode*code_sel / missing list ;
>     format code_sel $cdsel. ;
>     title1 "Checking Mapping Freqs off &file1: &tit1 Non-Crossover data" ;
> run ;
>
> proc summary data = &dsn1..&file1 missing ;
>     where cross = 'N' and code_sel ne '02' ;
>     class code_sel ;
>     var days_seg amt_bld amt_pd ;
>     format code_sel $cdsel. ;
>     output out = print1
>        sum =  ;
> run ;
>
> proc print data = print1 uniform ;
>     title1 "Print of &tit1 Non-Crossover accommodation codes" ;
> run ;
>
*******************************************************************************;
> proc freq data = &dsn1..&file1 ;
>     where cross = 'Y' ;
>     tables plancode*clmtype*procind*accmcode*code_sel / missing list ;
>     format code_sel $cdsel. ;
>     title1 "Checking Mapping Freqs off &file1: &tit1 Crossover data" ;
> run ;
>
> proc summary data = &dsn1..&file1 missing ;
>     where cross = 'Y' and code_sel ne '02' ;
>     class code_sel ;
>     var days_seg amt_bld amt_pd ;
>     format code_sel $cdsel. ;
>     output out = print1
>        sum =  ;
> run ;
>
> proc print data = print1 uniform ;
>     title1 "Print of &tit1 Crossover accommodation codes" ;
> run ;
>
*******************************************************************************;
>
>
> And the datasets are fairly good sized:
>
> NOTE: The data set SAS1.HIP01GAC has 6689825 observations and 50 variables.
> NOTE: The data set SAS1.HIP02GAC has 7699693 observations and 51 variables.
>
>   Volume in drive G is Local Data 2
>   Volume Serial Number is 3C8E-DDB0
>
>   Directory of G:\Data\Medi-Cal
>
> 03/13/2003  06:16p       1,522,336,768 hip01gac.sas7bdat
> 03/13/2003  06:16p         163,570,688 hip01gac.sas7bndx
> 03/13/2003  04:50p       1,802,191,872 hip02gac.sas7bdat
> 03/13/2003  04:50p         188,064,768 hip02gac.sas7bndx
>
>
>
> HTW,
>
> Out
>
>
>
> ------------------------------------------------------------
> William W. Viergever          Voice  : (916) 483-8398
> Viergever & Associates        Fax    : (916) 486-1488
> Sacramento, CA 95825          E-mail : wwvierg@attglobal.net
> ------------------------------------------------------------
>
>
>
>