AMD Phenom II 1090T getting too hot - is it dieing?

Toggle sidebar Toggle sidebar
D

DemiGoth

Honorable
Dec 25, 2013
14
0
10,510
I know, the 1090T is already ancient when it comes to age on CPUs, but so far it's been running very well in my PC with minimum to no heating. It topped at 65C most of the time when under 75-100% load.

It's been only recently that the temperatures have started to increase dramatically and now it's getting toward 75-80C.

I've kept my PC rather dust free (I live in a very old home so there's always more dust than a new one). I have recently replace the CPU fan and added an extra one (using CoolerMaster Hyper TX3 EVO) and exchanged the thermal paste as well, but it's still hitting those high temperatures :(

I'm wondering if these are the first signs of a CPU that's about to die? The CPU itself is not giving any errors at all though and aside from the heating up it's running very smooth. I have also checked my system for mallware (I do that on a regular basis, both with tools and manually) and there's none there.

My setup:
Asrock 990FX Extreme 3 (I know, shitty board - wish I knew that before I bought it)
GEiL 1600Mhz DDR3 2x8Gb paired RAM
MSI GTX970 Gaming
OCZ 750W Fatal1ty PSU
added an extra fan to the casing as well (still have 2 open, but none close to the CPU to remove heath).
 
Solution
S
I think its a simple case of spoiled thermal paste.

If you were running OK on CnQ befor. Running fan at max is not a solution rather a temporary work around., only reason it would go off on temp is because the thermal paste has dried out. Remove heatsink. Reapply paste. Reseat cooler. Maybe get an aftermarket cooler as well.

CPU does not just die unless you are overvolting, clocking or spilling water. Or do things out of ordinary. I had many PCs, many faults but never a spoiled CPU. I have had deterioration in temps, and always fixed by cleaning heatsink and fan and reapplying paate
Sort by date Sort by votes
R_1

R_1

Expert
Ambassador
Sep 8, 2015
8,440
632
49,290
have you stress tested the CPU? have you bench marked it to see if it has slowed down any?
Are/have you overclocking/overclocked?
Verified the proper voltages in the BIOS?
do you have access to any other computers? for swapping parts for testing.
 
Upvote 0 Downvote
D

DemiGoth

Honorable
Dec 25, 2013
14
0
10,510


I'm not overclocking and the BIOS voltages are approx correct (they always tend to flux a bit). CPU core (the most important one) varies between 1.176 and 1.448 depending on CPU speed used (which is between 795Mhz and 3624Mhz - latter one is the boosted speed)
There's no need for stresstesting, because a couple of years ago it went up to 65C when stresstesting it (with only 1 fan on the cooler as well). Now without really stressing it constantly (playing Black Desert online, which uses between 50% - 100% CPU) it's going to 75C easily. I bet that stresstesting the CPU now would hop to the temperature threshold and shut the PC down immediately.
Sadly, no parts here to swap with. I do have a 2nd PC with an AMD 965 BE on it, but that mobo can't handle over 95W (1090T is 140W).

[EDIT]
Perhaps worth to mention that the cores are now rated at 60C each and the CPU itself on 75C. Yet when I feel the air that's blowing out of the case through the fans (one right on top of the CPU), it feels like a warm wind (40C-ish), rather than very hot wind.
I also touched the heathsink left of the CPU (northbridge, right) before I applied the new paste, and it did feel warm to the touch, but I didn't burn my finger on it. That'd imply that the thing is not 75C. Same goes for the CPU heathsink as well (didn't touch the copper parts though).
 
Upvote 0 Downvote
R_1

R_1

Expert
Ambassador
Sep 8, 2015
8,440
632
49,290


but your current system can handle the 965. try swapping the CPU's and see if the 965 is getting hotter than it should, this may be an indication of a power problem, either motherboard or PSU. if the 965 works just as it should then the CPU (1090) is suspect.
you should also have a power supply you can swap for testing so you are in better shape than most.
 
Upvote 0 Downvote
D

DemiGoth

Honorable
Dec 25, 2013
14
0
10,510


Sadly, the PC which has the 965 only has a 300W PSU. It's my Linux machine that I'm using for development tests (hobby programmer here) only.

Then I thought of my son's PSU, which is a 620W one, and should be able to hold the GTX970. And while thinking about that, I remembered that his motherboard also can handle my 1090T. So I think I'll drop the 1090T in his PC tomorrow (he's away for holiday anyway) and test it there. If the temperatures get this high there, it's save to say it's the CPU. If not, I'll pull his PSU and replace it with mine and see if the problem remains. If it then still remains it's save to say that the (shitty) Asrock board is the heart of the problem.

...that's the right way to go I think, right..?
 
Upvote 0 Downvote
D

DemiGoth

Honorable
Dec 25, 2013
14
0
10,510


Took a peek in my BIOS and changed some in fan control. I also found a setting on AUTO which relates to AM3/AM3+ toggling. Though I figure the BIOS should do that correctly, I decided to set it hard to AM3. After that I took a burn test (for 12 minutes) and the results are 'normal' again...

Core temperature were constant on 50C and raised to 51C at the 10th minute
CPU temperature was constant on 67C and raised to 68C on the 10th minute

After the burn test stopped the cored dropped down to 30-ish within a couple of seconds and now (2-3 minutes later) they are back on 25C. CPU temperature kinda dito story, it dropped down to 55C within a couple of seconds and is now back to 42C. These values are the ones I was used to with my previous ASUS board.

I think it's been one of the BIOS settings I fiddled with, but will keep a close eye on the temperature for the coming days. If they start to behave strange once again I'll start on the swapping with my son's PC to test it further...
 
Upvote 0 Downvote
D

DemiGoth

Honorable
Dec 25, 2013
14
0
10,510
After having changed some of the BIOS settings, including turning off the one where the auto throttle up to x18 speed is handled (3.6Ghz), the CPU seems stable again. Somehow the throttling AND Cool & Quiet were the ones causing the heating problem I've experienced.

The Cool & Quiet is not turned off and while the fans run to the max, I'm using silent ones and don't hear them as noisy (yes they do make noise at full speed, but they're not loud - my old GTX660 made more noise when the fans ran to the max :D )
The throttle is something different. Now that the CPU stays on 3.2Ghz (as it did with my ASUS board) the overheating seems gone. I think that my old CPU is not meant for (auto) overclocking and with the age of the CPU (6 years already) I think it's best to keep it at it's intended speed instead of throttling it up to 3.6Ghz and risk a shorter lifespan (though the CPU is already over it's 'expiration date').
 
Upvote 0 Downvote
S

sarwar_r87

Distinguished
Mar 28, 2008
837
0
19,060
I think its a simple case of spoiled thermal paste.

If you were running OK on CnQ befor. Running fan at max is not a solution rather a temporary work around., only reason it would go off on temp is because the thermal paste has dried out. Remove heatsink. Reapply paste. Reseat cooler. Maybe get an aftermarket cooler as well.

CPU does not just die unless you are overvolting, clocking or spilling water. Or do things out of ordinary. I had many PCs, many faults but never a spoiled CPU. I have had deterioration in temps, and always fixed by cleaning heatsink and fan and reapplying paate
 
Upvote 0 Downvote
Solution
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom