90's on core i5 cpu temp. is that bad?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

grkM3

Golden Member
Jul 29, 2011
1,407
0
0
I apologize in advance but I almost wish your CPU would fry so you would realize how crazy you sound.

Seriously, you think nothing is wrong with running 90+C 24/7 just because of auto throttle?

no you need to start reading more carefully and stop saying things I never say.

If you took the time to read the person I quoted he said that 80c will degrade a chip.

A STOCK 2500 WILL HIT OVER 80C RUNNING LINX....do you understand?for some reason you think im running 98c for 24/7

all im saying is a stock cooler,stock clocked chip will hit over 80c running full load.

Why dont you do this,run 10 min of linx set to 8 threads and 4gb ram and post up your cpu temps,make sure you have HT also.

I cant wait to see how cold your chip is running,I just ran 32mil pi,winamp and 3dmark 03 all at the same time and my chip never saw over 73c with the same exact settings I had when it throttle running intel burn test
 
Last edited:

grkM3

Golden Member
Jul 29, 2011
1,407
0
0
I just tried intel burn test for about 5 minutes or so - i just set the stress level to max and let it run. It completed one run and was in it's 2nd run when i stopped it. highest temp i saw was 84.

but using these to test the cpu seems a little unrealistic, unless there are applications out there that might stress the cpu in the same way. I'd rather stress the cpu under what it might normally get stressed with - gaming, video editing etc and see what the temp reaches.

that is what intel uses to stess the cpu,see you hit 84c in under 5 min,people run that program for hours over at the other forums to show there overclocks are stable.

Is your case closed or open?
 

mrSHEiK124

Lifer
Mar 6, 2004
11,488
2
0
The Corsair A50 is on clearance at a lot of Best Buys for $20. Save yourself some headache. Ignore the crazies, running a chip above 80C (unless it's Fermi) for an extended period of time is NOT A GOOD IDEA. Your room doesn't feel warmer?
 

Rifter

Lifer
Oct 9, 1999
11,522
751
126
The Corsair A50 is on clearance at a lot of Best Buys for $20. Save yourself some headache. Ignore the crazies, running a chip above 80C (unless it's Fermi) for an extended period of time is NOT A GOOD IDEA. Your room doesn't feel warmer?

its actually not so bad, with stock cooling the original i7 as well as SB i7's will hit 80c under stress. I wouldnt want to run it 24/7 at 95c but hitting the 80's is no big deal.
 

toyota

Lifer
Apr 15, 2001
12,957
1
0
The Corsair A50 is on clearance at a lot of Best Buys for $20. Save yourself some headache. Ignore the crazies, running a chip above 80C (unless it's Fermi) for an extended period of time is NOT A GOOD IDEA. Your room doesn't feel warmer?
whether the chip is 70c or 90C under load its still using the same amount of wattage. your room does not really get any hotter because of a less effective cooler.
 

merk

Senior member
May 29, 2003
471
9
91
that is what intel uses to stess the cpu,see you hit 84c in under 5 min,people run that program for hours over at the other forums to show there overclocks are stable.

Is your case closed or open?
i dont know why this is devolving into an argument of what temp the cpu can/should run under.

Personally, i think running a cpu in the 90's is too high. running it in the low 80's seems pretty high to me. But since i only hit mid 80's when using a program designed specifically to stress the cpu, i'm not too worried. Since i wont be normally running programs for the sole purpose of using the cpu.

Runng handbreak, which is what i would consider a real-world stress test, the cpu hit the high 60's or just tipped into 70. So i think the stock cpu cooler is doing it's job since i probably wont run into very many situations where the cpu is going to be maxed out like that non-stop, other then using something like handbrake.

From what others have said on here, the argument doesn't seem to be if it's stable at those high temps, but how long your cpu will last over it's lifetime of use before those high temps cause it to fail. I'm guessing running at 70c for a couple of hours once in a while is just fine.
 

Ben90

Platinum Member
Jun 14, 2009
2,866
3
0
whether the chip is 70c or 90C under load its still using the same amount of wattage. your room does not really get any hotter because of a less effective cooler.
I don't mean any hostilities from this comment, because I'm sure you already know. Technically a hotter chip will use SLIGHTLY more power for real world temps.

I'd argue it matters about as much as running your chip right against the TJMax in most situations. Doesn't really matter enough to worry about it.

blackened23 said:
I apologize in advance but I almost wish your CPU would fry so you would realize how crazy you sound.

Seriously, you think nothing is wrong with running 90+C 24/7 just because of auto throttle?
This seems to be a reoccurring theme on Anandtech recently. We have two sides to this problem. Side one believes that temps in the 80-100*C range are absolutely horrible for our processors because they, just well are. Side two believes that a Multibillion company wants to reduce things like warranty returns and create a positive brand image.

Its not like Intel knows how to test their chips and keep them running safe at all times. No, This is Anandtech 2011! We know better then them. If your processor hits 80.1*C it blows up instantly.
 

mikeymikec

Lifer
May 19, 2011
20,427
15,161
136
I think in the long term 80 - 90C CPU temps are bad for the system as a whole. While the CPU has been tested to handle a wide range of temperatures, it doesn't mean that nearby components have as well.

Otherwise, benchmark the CPU and/or play games until your heart is content.
 

LOL_Wut_Axel

Diamond Member
Mar 26, 2011
4,310
8
81
I don't mean any hostilities from this comment, because I'm sure you already know. Technically a hotter chip will use SLIGHTLY more power for real world temps.

I'd argue it matters about as much as running your chip right against the TJMax in most situations. Doesn't really matter enough to worry about it.


This seems to be a reoccurring theme on Anandtech recently. We have two sides to this problem. Side one believes that temps in the 80-100*C range are absolutely horrible for our processors because they, just well are. Side two believes that a Multibillion company wants to reduce things like warranty returns and create a positive brand image.

Its not like Intel knows how to test their chips and keep them running safe at all times. No, This is Anandtech 2011! We know better then them. If your processor hits 80.1*C it blows up instantly.

NO ONE in this thread has said that. Stop putting words into our mouth.

They don't. All they list is the max temp the CPU should run at any given time. For long term use more than 80C will cause degradation.

Both high temperatures and high voltage degrade a CPU. A combination of both will mean imminent degradation most of the time in less than a year, while only high temperatures would degrade the CPU in a much longer time frame of 5+ years.
 

Ben90

Platinum Member
Jun 14, 2009
2,866
3
0
Unfortunately the Arrhenius Equation doesn't quite agree with your specific timescale. In much the same way as it takes 4+ minutes to run a mile, it basically takes every one of us more than 4 minutes to run that distance. On the extreme side, one of us may achieve a 3:59 time, and a CPU may die within its warranty period because of heat related problems.

Extrapolating someone who worked in process manufacturing's indirect response on the matter, the CPU in question should theoretically be designed to run 40 years at 80*C with temperature and voltage being our only catalysts in the system. 12 years as a realistic minimum bound and 3 as an absolute bound.

40 years applying the 10*2*2 SOP design requirements + Arrhenius.
12 years applying the 3*2*2 warranty + Arrhenius
3 years applying the 3 year warrenty.
 

blackened23

Diamond Member
Jul 26, 2011
8,548
2
0
This seems to be a reoccurring theme on Anandtech recently. We have two sides to this problem. Side one believes that temps in the 80-100*C range are absolutely horrible for our processors because they, just well are. Side two believes that a Multibillion company wants to reduce things like warranty returns and create a positive brand image.

Its not like Intel knows how to test their chips and keep them running safe at all times. No, This is Anandtech 2011! We know better then them. If your processor hits 80.1*C it blows up instantly.

Nah, nobody ever said that. I've been overclocking for a while (ever since the P55C vanilla pentium days) and i've had a few chips die, and they all died after 1+ year of continuous overclocking. While I used to be super aggressive about overclocking in the past, going shit crazy about overclocking everything and anything in my system, i'm a bit more cautious now. Anyway, running 80+c continuously will harm the chip. It doesn't happen overnight, it might not happen in 6 months, but in my personal experience it will slowly degrade the chip over time. I don't have any scientific proof but i've had chips die after long periods of rock solid stable overclocking, mostly after 12+ months of use. Specifically I remember having a 300a overclocked super high that died after 11 months of continuous use, with the PC running 24/7, and the CPU did run hot. Could it have been due to something else? Maybe, but I doubt it. I also had a lynnfield chip die with a super aggressive overclock, but another one is still going. So I know that I will err on the side of caution, thats me though. I had the same attitude that nothing could ever go wrong before too, I thought 1 night of continuous stress testing passed meant that my chip was invincible.

Then again CPU's and motherboards are pretty cheap nowadays, so whatevs.
 
Last edited:

grkM3

Golden Member
Jul 29, 2011
1,407
0
0
Your chips died from voltage spikes and too much current.Degration comes from phisically burning out the transistors with too much current along with high clocks.

you can degrade a 2600k on liquid nitrogen running -30c at 5.8ghz by jamming 1.65 volts into it.

I also want to note that my 2600k has the gpu turned off so it will hit even higher temps with it on and using the stock intel appoved heatsink.

For some reason this board thinks 80c at stock clocks and stock voltage is bad for the chip.Well then why didnt intel use a bigger heat sink then?Do you guys know more than intel?

Im still waiting for your linx runs blackend,you seem to have something against me on everyone of my posts.I am willing to bet that your very own chip is in the same heat range as what you preach as being to hot and bad for the chip.
 
Last edited:

grkM3

Golden Member
Jul 29, 2011
1,407
0
0
i dont know why this is devolving into an argument of what temp the cpu can/should run under.

Personally, i think running a cpu in the 90's is too high. running it in the low 80's seems pretty high to me. But since i only hit mid 80's when using a program designed specifically to stress the cpu, i'm not too worried. Since i wont be normally running programs for the sole purpose of using the cpu.

Runng handbreak, which is what i would consider a real-world stress test, the cpu hit the high 60's or just tipped into 70. So i think the stock cpu cooler is doing it's job since i probably wont run into very many situations where the cpu is going to be maxed out like that non-stop, other then using something like handbrake.

From what others have said on here, the argument doesn't seem to be if it's stable at those high temps, but how long your cpu will last over it's lifetime of use before those high temps cause it to fail. I'm guessing running at 70c for a couple of hours once in a while is just fine.

its becasue some think 80c will degrade your chip...btw your cpu should run at 97c for 3 years using that burn test,you mother board will die way before your chip does though from stressing.It might last 5-10 days max before something gives up but thats why we buy better boards with better components to sustain longer loads and clocks.
 
Last edited:

merk

Senior member
May 29, 2003
471
9
91
its becasue some think 80c will degrade your chip...btw your cpu should run at 97c for 3 years using that burn test,you mother board will die way before your chip does though from stressing.It might last 5-10 days max before something gives up but thats why we buy better boards with better components to sustain longer loads and clocks.
for me it's irrelevant. Like i said, the only thing so far that really seems to stress the cpu is a program whose sole purpose is to stress the cpu. If a real world app like handbrake only pushes the temp to 70c then i think i'm just fine. I doubt the cpu or the mobo is going to fail at those temps, especially since i'll only be running at those temps for relatively short periods of time.
 

Rifter

Lifer
Oct 9, 1999
11,522
751
126
for me it's irrelevant. Like i said, the only thing so far that really seems to stress the cpu is a program whose sole purpose is to stress the cpu. If a real world app like handbrake only pushes the temp to 70c then i think i'm just fine. I doubt the cpu or the mobo is going to fail at those temps, especially since i'll only be running at those temps for relatively short periods of time.

load up civ 5 with a large map and then play till you are half way through the game. That heats my CPU up more than prime95 or OCCT CPU burn test.

There are plenty of apps out there that will stress your CPU to the max, this is why people that overclock will use programs like prime95 and linx to test the overclock because while rare there are real worl situations that will push your CPU to the limit.
 

merk

Senior member
May 29, 2003
471
9
91
Will the demo version stress it? if so i'll go install it and give it a try just to see what temps i hit.
 

Rifter

Lifer
Oct 9, 1999
11,522
751
126
Will the demo version stress it? if so i'll go install it and give it a try just to see what temps i hit.

i dunno i never played the demo. All i know is in a large map with 8 players when it gets to the point where the AI turns are taking a min or two at a time it works your CPU harder than any other game i have played. I was also playing with maxed graphics settings including max AA and tesselation i dont know if thats a factor or not.
 

merk

Senior member
May 29, 2003
471
9
91
ah so it was a multiplayer game? I'm guessing the demo probably doesn't support multi-player - plus i'm not sure i want to play long enough to actually go through a whole multiplayer map.
 

Rifter

Lifer
Oct 9, 1999
11,522
751
126
ah so it was a multiplayer game? I'm guessing the demo probably doesn't support multi-player - plus i'm not sure i want to play long enough to actually go through a whole multiplayer map.

yeah was multi, was playing with a few friends and a few AI opponents. Gets really stressful on the sytem when you need to wait for the AI to figure out what to do in the late game.