mersenneforum.org  

Go Back   mersenneforum.org > New To GIMPS? Start Here! > Information & Answers

Reply
 
Thread Tools
Old 2020-11-21, 14:17   #23
wagner85
 
Aug 2020

268 Posts
Default

Quote:
Originally Posted by axn View Post
This is also consistent with cooling issue. If issue is with PSU, that would result in system instability / crashes. This looks more like CPU is throttling down due to heat.

watch "grep 'cpu MHz' /proc/cpuinfo" will show if CPU is throttling. But I have no solution if that is the case (apart from reinstalling / replacing cooler)

It does not seem to be thermal throttle. These CPUs are watercooled.check the pics for data.
The temps were never so low. Unless another region of the MB that does not have sensors is overheating. The rigs I have are open 1u servers. The old cooling system cooled everything around the MB. This new cooling system cools just de CPUs.

The cpus work at the max they can, without turboboost, generating 16 ms/itr
I acknowledge the need of changing the psu. However, I want to be 100% sure before spending the money.
Today I will remove one of the CPUs and run the test again.
If the PSU is the limiting factor for 2 cpus, using just one will not draw as much energy as two do. So I would expect the single cpu delivering around 7ms/itr.
Attached Files
File Type: pdf Cpu Throttle test _compresseda.pdf (1.50 MB, 6 views)
wagner85 is offline   Reply With Quote
Old 2020-11-21, 15:05   #24
wagner85
 
Aug 2020

1616 Posts
Default VBCurtis, Axn, et al. Thank you

Issue fixed!( partially). Need to upgrade the psu.

Thermal throttle was the issue.
not on the CPUs, but Somewhere in the MB.
I placed my hands in some of the components and noticed their temps were quite high.
For now I left a big cooler blowing air on it. And it is working greatly.
However, the power consumption is around 507W.
I will temporary fix the issue by reducing the cores I use.
Thanks again guys.



#2 Nov 21 08:42] Iteration: 24904700 / 108897011 [22.86%], ms/iter: 7.363, ETA: 7d 03:47
[Worker #1 Nov 21 08:42] Iteration: 74324800 / 112583063 [66.01%], ms/iter: 7.450, ETA: 3d 07:10
[Worker #2 Nov 21 08:42] Iteration: 24904800 / 108897011 [22.87%], ms/iter: 7.432, ETA: 7d 05:24
[Worker #1 Nov 21 08:42] Iteration: 74324900 / 112583063 [66.01%], ms/iter: 7.491, ETA: 3d 07:36
[Worker #2 Nov 21 08:42] Iteration: 24904900 / 108897011 [22.87%], ms/iter: 7.504, ETA: 7d 07:04
[Worker #1 Nov 21 08:42] Iteration: 74325000 / 112583063 [66.01%], ms/iter: 7.491, ETA: 3d 07:36
[Worker #2 Nov 21 08:42] Iteration: 24905000 / 108897011 [22.87%], ms/iter: 7.364, ETA: 7d 03:4
wagner85 is offline   Reply With Quote
Old 2020-11-21, 16:05   #25
axn
 
axn's Avatar
 
Jun 2003

17·281 Posts
Default

axn is offline   Reply With Quote
Old 2020-11-22, 06:51   #26
Mark Rose
 
Mark Rose's Avatar
 
"/X\(‘-‘)/X\"
Jan 2013

2×31×47 Posts
Default

The rating of a PSU is how much it can supply, not how much it draws. A 500 watt power supply drawing 507 watts is running under its rated capacity (basically no power supply is 98% efficient). That being said, it is harder on the components and it will wear out faster. It's likely the system was not originally supplied with such high wattage CPUs, and that the motherboard was designed to have airflow over it that the water cooling isn't supplying.
Mark Rose is offline   Reply With Quote
Old 2020-11-23, 02:56   #27
wagner85
 
Aug 2020

2·11 Posts
Default

Quote:
Originally Posted by Mark Rose View Post
The rating of a PSU is how much it can supply, not how much it draws. A 500 watt power supply drawing 507 watts is running under its rated capacity (basically no power supply is 98% efficient). That being said, it is harder on the components and it will wear out faster. It's likely the system was not originally supplied with such high wattage CPUs, and that the motherboard was designed to have airflow over it that the water cooling isn't supplying.
you are completely correct.
These are 1u servers I run p95!
The fans were too noisy. around 60 dB. too much for me.
so I decided to replaced all the fans.

Last fiddled with by wagner85 on 2020-11-23 at 02:57
wagner85 is offline   Reply With Quote
Old 2020-11-24, 18:35   #28
wagner85
 
Aug 2020

101102 Posts
w00t

Quote:
Originally Posted by VBCurtis View Post
If you're drawing 530W from a 500W power supply, your hardware will not last long. If you're somewhat lucky, the power supply will fail. If you're less lucky, it'll fail in a way that blows up the motherboard, or a CPU, or maybe a smoky mess when it goes...

Get a new power supply. One big enough to power your machine at less than 80% load- so, 700W minimum.
I ended up buying a power supply on Newegg. Close to 50.00 with rebate! 69 full price.

Enermax CyberBron ECB700EWT 700W ATX12V 80 PLUS BRONZE Certified Non-Modular, 5 Year Warranty, Active PFC Power Supply.


However, I believe my 500W PSU handles the server well.
I was under the impression I was pulling out of the power supply more than it could give.
However, reading little bit further I checked that my 500 W PSU efficiency is 87% at 100% usage.
So to be able to supply 500W to the server internal components, it needs to widraw from the wall 500W/0.87= 568 W.
So according to what I read my PSU widraws 568W from the wall to supply my server with 500W at 87% efficiency.

Source:
https://www.velocitymicro.com/blog/w...-it-important/

Could you all please double check my assessment ?

Because if that is true I will keep using my current PSU and probably build something else beneficial to gimps with this power supply.
wagner85 is offline   Reply With Quote
Reply

Thread Tools


Similar Threads
Thread Thread Starter Forum Replies Last Post
Lower Lip R.D. Silverman Soap Box 10 2013-07-30 14:50
lower bounds on incomplete factorizations J.F. Factoring 3 2008-06-14 18:58
Lower bounds for odd multiperfect numbers. jchein1 Math 7 2006-11-26 13:29
How do I lower Prime95's CPU Usage? Unregistered Software 21 2003-12-13 15:26
Where do I find undoc.txt to lower the usage? Unregistered Software 3 2003-12-12 21:47

All times are UTC. The time now is 19:06.

Thu Nov 26 19:06:55 UTC 2020 up 77 days, 16:17, 3 users, load averages: 1.61, 1.45, 1.47

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.