[Brahms-dev-l] Fw: Your jobs on rcas0056

From: Flemming Videbaek <videbaek_at_bnl.gov>
Date: Wed, 15 Nov 2006 19:34:05 -0500
a bit more info - be aware tough it might not be easy to figure out what jobs have been run on this machine

flemming

--------------------------------------------
Flemming Videbaek
Physics Department 
Bldg 510-D
Brookhaven National Laboratory
Upton, NY11973

phone: 631-344-4106
cell:       631-681-1596
fax:        631-344-1334
e-mail: videbaek @ bnl gov
----- Original Message ----- 
From: "Chris Hollowell" <hollowec_at_bnl.gov>
To: "Flemming Videbaek" <videbaek_at_bnl.gov>
Sent: Wednesday, November 15, 2006 6:41 PM
Subject: Re: Your jobs on rcas0056


>> Thnaks for the information;
>> Overall we have been in reasonable luck,since some but not too much 
>> analysis had been done on
>> this machine since Friday. Do you have some more details on the errors e.g 
>> would 'normal' multiply,
>> divide be ok?
> 
> Integer division/multiplication likely worked properly.  Any
> floating point operations should be suspect, however.  The simple 
> act of placing a constant float value into an FPU register 
> produced incorrect results: i.e., 3.1415... became 4.000000175...
> 
> 
> Chris
>
_______________________________________________
Brahms-dev-l mailing list
Brahms-dev-l_at_lists.bnl.gov
http://lists.bnl.gov/mailman/listinfo/brahms-dev-l
Received on Wed Nov 15 2006 - 19:34:53 EST

This archive was generated by hypermail 2.2.0 : Wed Nov 15 2006 - 19:35:14 EST