What is the impact of running the CPU at 100% for long periods at a time?

15

4

Possible Duplicate:
Is it ok to have 100% CPU usage constantly

I recently upgraded computers and with this upgrade I decided to try Seti@Home to run in the background. I was watching the CPU and monitoring how it works and when idle the CPU is running at 100% (due to Seti). But does drop down when I start to do things with the computer.

With the CPU running at 100% quite often can it damage the CPU over time? I want this computer to last as long as possible. The temps for the CPU when idling is around 100 degrees Fahrenheit (37 degrees Celsius) and when at 100% averages 130-140 degrees Fahrenheit. (54 - 60 degrees Celsius) I am running an AMD FX 4100 CPU

The question is this: What is the impact of running the CPU at 100% for long periods at a time? Does this degrade the life of the CPU or other components like overclocking can/does?

EDIT- The original temps posed in this question where very low and were a faulty reading from a program. The correct temps are now in the question.

L84

Posted 2012-06-01T06:23:01.297

Reputation: 3 403

Question was closed 2012-06-01T12:20:03.837

2Your idle temperatures are below room temperature here (it's also winter and cold). Your load temperatures are only slightly higher than my idle temperatures, and about the same when it's summer here. That's nowhere near the max limits of any consumer CPU. – Bob – 2012-06-01T06:26:35.587

1If you're worried about over-working your CPU, then stop running SETI. What's the actual problem you're encountering? Are you asking if that's "Too hot" for that processor? Or do you just want to discuss the potential impact of using a CPU fully for long periods of time? If it's the later, than this is the wrong place to ask. – Ƭᴇcʜιᴇ007 – 2012-06-01T06:27:06.020

@techie007 - Why is it wrong to ask if running the CPU for extended periods of time off topic for SuperUser? That is my question, what is the impact of running the CPU at 100% for extended periods? Also not discuss ask what is the impact. I assume there are studies etc. And I will stop running SETI but if research has shown that there is not much inpact then I will not. – L84 – 2012-06-01T06:31:17.200

1Those temperatures look more typcial for celsius units. Are you sure it's F? – Keith – 2012-06-01T07:13:54.090

Also depends on the quality of your system. Cheap systems will overheat and shut themselves down. – Keith – 2012-06-01T07:14:27.740

@Keith - Yes that is correct. The readings are in Fahrenheit. I have edited the question showing the Celsius conversions. I will say once I saw the temp I was shocked. The last computer at idle ran at 114 degree Fahrenheit (it was also 9 years old Pentium 4 => ). This is the coolest machine I have ever seen. As a side note my room while not hot is not the coolest either and it maintains these temps. – L84 – 2012-06-01T07:37:48.827

2Folding@Home client on a laptop: Is it safe for long periods? – Sathyajith Bhat – 2012-06-01T08:29:31.970

1Thats insanely cold. What's your ambient temperatures like? Your fully loaded temperature is what my c2d runs at normally, and its been going fine since 07. Also, what are you using to monitor your temperatures? – Journeyman Geek – 2012-06-01T09:26:58.743

@JourneymanGeek - Core Temp is the program I was using to monitor the temps. After your (and the above) comments I installed another program (Open Hardware Monitor) and it was giving me temps around 100 degrees Fahrenheit at idle and at max load it would range from 120-130 degrees Fahrenheit. Do you have a program you recommend that I can try to see if I can confirm one or the other temp? – L84 – 2012-06-01T17:23:19.777

@JourneymanGeek - As a follow up here I checked in the BIOS and the idle temp was around 100 (F) I would say that Core Temp was way off, I excepted a little error but not that much. I have corrected the temps in the question even though it was closed. – L84 – 2012-06-02T14:03:55.763

Answers

4

The question is this: What is the impact of running the CPU at 100% for long periods at a time? Does this degrade the life of the CPU or other components like overclocking can/does?

To answer your question, yes.. running a cpu at full capacity for extended periods will shorten it's lifespan. A combination of heat and flow of electricity through the chip causes changes to the components within the chip at the atomic level, known as electromigration. Here's a brief (and very simplified) explanation:

This is when metal atoms wander into the dividing layers on a microprocessor. It is caused by the combination of electricity and heat. Processors are designed to run within certain heat and electrical specifications, and if run at higher heat and/or electrical specifications, electromigration may occur. If this occurs to a great degree and enough metal atoms wander off of the lines in a processor, they may permanently ruin the processor by thinning a connection so that it does not work effectively, or even making an electrical connection where one is not intended to be. Overclocking and raising voltage supplied to a processor increases the risk of electromigration.

http://www.geek.com/glossary/E/electromigration/

A more in depth explanation can be found here: http://www.csl.mete.metu.edu.tr/Electromigration/emig.htm

So taking this information into consideration we can deduce that more heat and more volts cause an increased level of degradation due to electromigration. Running the CPU at 100% means it will receive more volts and therefore more heat (for sustained periods), ultimately contributing to the shortened lifespan (literally speaking, without taking into account how long and how often these periods of activity occur). Of course, the point is almost moot because as you state, your temperatures are good and well within thermal limits so it would take a long time for the processor to degrade to the point of failure - i'd be more concerned about the electricity bill!

With regards to overclocking, to attain higher clocks - more volts are required at the expense of increased thermal output. Taking into consideration the information above, pushing more volts through a processor will cause the chip to degrade a lot faster (timescale obviously depends on the amount of volts and how the processor is used).

Hope this information helps answer your question..

Dean

Posted 2012-06-01T06:23:01.297

Reputation: 674

3-1. The text that you quote is right: "Processors are designed to run within certain heat specifications; if run at higher heat specifications, electromigration may occur". That means that the correct answer is NO, there is no harm done from running at 40 degrees C. That is entirely within the designed heat specifications. – MSalters – 2012-06-01T11:22:42.567

@MSalters - Read the answer again.... 'Of course, the point is almost moot because as you state, your temperatures are good and well within thermal limits so it would take a long time for the processor to degrade to the point of failure - i'd be more concerned about the electricity bill!' – Dean – 2012-06-01T11:26:27.973

You're drawing the assumption that the safe heat level (below which the CPU isn't damaged) lies at less than 100% nominal CPU load. This is not supported by your quotes. It in fact suggests that there's no damage done until you overclock, and only then will the damage increase with voltage/heat. – MSalters – 2012-06-01T11:31:59.027

@MSalters - My assumption in this case, is that running the CPU at maximum load for extended periods raises the the overall thermal output, correct? Whether or not that is inside or outside the heat specification of the chip is irrelevant. Lifespan of the CPU would be shorter running constantly at 50C that it would at 40C would it not? Although, I do see you point in the sense that quote suggests electromigration may only occur once outside the heat specifications.. – Dean – 2012-06-01T11:49:46.103

3@Dean - There is no evidence that running a CPU at 100% load for 5 years would shorten its life from say 6 years. Since it is unrealistic to perform a scientific to prove your statements ( considering every single CPU dye is unqiue ) just don't make the claim. Please provide evidence of your claims. If you run the CPU within the specifications the lifespan in theory would not be reduced. It is not possible to estimate when a CPU will stop working. – Ramhound – 2012-06-01T12:04:17.477

1@Dean: Essentially, heat (thermal energy) and voltage (electrical energy) cause electromigration by physically moving atoms outside their crystal position. If the available energy isn't high enough, the atoms will just bounce back to their initial lattice position. – MSalters – 2012-06-01T12:17:14.603

you are both right.. its a shame the question is closed, i enjoy some intelligent conversation (even if i am wrong lol) – Dean – 2012-06-01T13:07:23.947