mersenneforum.org  

Go Back   mersenneforum.org > Great Internet Mersenne Prime Search > Software

Reply
 
Thread Tools
Old 2006-04-26, 03:05   #1
PhilF
 
PhilF's Avatar
 
Feb 2005
Colorado

10000010112 Posts
Default Unusual error - comments George?

I received a SUM(INPUTS) != SUM(OUTPUTS) error.

The unusual part is that a got a "Disregard last error. Result is reproducible and thus not a hardware problem." I thought reproducible errors only happened on round off errors when testing an exponent near the FFT size limit. This exponent (34685927) is not near an upper FFT boundary limit.

Thoughts? Has anyone ever had a SUM(INPUTS) != SUM(OUTPUTS) error that was reproducible?

Additional info: The two sums are very close:

9.995465689403947e+16
9.995465689403128e+16
PhilF is offline   Reply With Quote
Old 2006-04-26, 17:34   #2
axn
 
axn's Avatar
 
Jun 2003

22×11×107 Posts
Default

Assuming that you are using a P4, did it use the 1792K FFT or 2048K FFT? I am guessing that it is the former which would explain the error.

Just a SWAG
axn is offline   Reply With Quote
Old 2006-04-26, 20:19   #3
Prime95
P90 years forever!
 
Prime95's Avatar
 
Aug 2002
Yeehaw, FL

7·1,019 Posts
Default

This "error" can happen with any FFT size, any exponent. It is an error check that compares two floating point values for "equality". Since it is impossible to compare floats for exactly, it makes sure the difference is "small". Small is computed based on some historical data and adjusted so that false error reports will be rare.

It looks like you just had an iteration where the difference was several standard deviations greater than normal.
Prime95 is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Unusual folly fivemack Lounge 1 2013-04-25 16:32
Query for George about ERROR 3 garo PrimeNet 17 2005-10-18 21:01
Unusual differentiation mfgoode Math 12 2004-08-25 16:33
unusual sequence tom11784 Puzzles 5 2004-05-20 12:35
Unusual Cruncher Deaths... E_tron Lounge 6 2003-11-18 23:05

All times are UTC. The time now is 06:34.

Thu Oct 22 06:34:11 UTC 2020 up 42 days, 3:45, 0 users, load averages: 1.54, 1.40, 1.29

Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2020, Jelsoft Enterprises Ltd.

This forum has received and complied with 0 (zero) government requests for information.

Permission is granted to copy, distribute and/or modify this document under the terms of the GNU Free Documentation License, Version 1.2 or any later version published by the Free Software Foundation.
A copy of the license is included in the FAQ.