mersenneforum.org  

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

Reply
 
Thread Tools
Old 2020-05-08, 06:12   #1
jiangv
 
Jul 2019

13 Posts
Default Why large fft test given so much stress to cpu?

recently I use latest version of mprime 95 to test my mem kits, I found that if I choose large fft(test mem controller and ram), the core temp could went up to 80 degrees, almost the same as I use small fft(L1-L3, max temp) to stress cpu only.

so why it would be some much stress given to cpu when I stressing mem?

Last fiddled with by jiangv on 2020-05-08 at 06:25
jiangv is offline   Reply With Quote
Old 2020-05-09, 19:42   #2
ewmayer
2ω=0
 
ewmayer's Avatar
 
Sep 2002
Rep├║blica de California

2C1616 Posts
Default

Really good FFT implementations will stress both the CPU and the entire memory subsystem.

If your CPU has more than 4 physical cores you may well see less CPU stress for big-FFT since the memory typically can't keep up with the data demands of all the cores, but I'll defer to the owners of such systems to confirm or deny that.
ewmayer is online now   Reply With Quote
Old 2020-05-11, 13:07   #3
jiangv
 
Jul 2019

11012 Posts
Default

Quote:
Originally Posted by ewmayer View Post
Really good FFT implementations will stress both the CPU and the entire memory subsystem.

If your CPU has more than 4 physical cores you may well see less CPU stress for big-FFT since the memory typically can't keep up with the data demands of all the cores, but I'll defer to the owners of such systems to confirm or deny that.
Thanks, my pc has 18 cores and I do see temps at normal reading about 60-70 degrees, But still dont understand why large fft will cause such stress as much as small fft
jiangv is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Please provide me really large numbers to test manually bosons Factoring 29 2019-11-12 13:58
P95 PrimeNet causes BSOD; small FFTs, large FFTs, and blend test don't KarateF22 PrimeNet 16 2013-10-28 00:34
Does a large P-1 test eliminate need for ECM tests UberNumberGeek Factoring 53 2009-07-11 08:24
Small prime test divisor efficiency strategy fenderbender Math 5 2007-07-18 18:39
blend / small FFT torture test on linux ? lukasz Linux 1 2007-01-09 16:16

All times are UTC. The time now is 01:45.

Wed Jul 8 01:45:00 UTC 2020 up 104 days, 23:18, 0 users, load averages: 1.43, 1.50, 1.54

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.