Serious overheating i5 3570k (100C), future options?

Toggle sidebar Toggle sidebar
P

papality

Honorable
Dec 1, 2012
74
0
10,640
So my CPU had been running at 4 GHz since I built in late 2012, air cooled Hyper 212 EVO, and is currently/was until very recently having major overheating issues, going all the way up to 100C. Never had issues with heat until this. It had been doing it off and on for a couple of weeks but I always managed to stop whatever it was doing and drop it quickly. So now I've gone into my BIOS and dialed the CPU back down to stock, rearranged my case Corsair 500R for better airflow by taking out unused HDD bay and DVD player (literally haven't used a disk in it in years), cleaned out a ton of dust from all over.

Ran a stress test in CPUID and it was still polling in at 92-99 C. I've been thinking about doing a new build anyway for a little while now and with Ryzen out I'm all but sure to start soon, just a matter of budgeting and all. I have a Corsair H60 and new thermal paste on the way, partly because I'd rather give my first shot at liquid cooling on something that's not a massive, brand-new investment.

So my main question is: given these temp concerns, should I accelerate my timeline for a new build? Is there any hope that the H60 could cure what ails me and buy me some more time?

CPU: i5 3570k w/Hyper 212 EVO
Mobo: Asus P8Z77-V LK
RAM: 24 GB 1600 MHz Patriot Intel Extreme Masters
OS SSD: 128 GB Samsung 840 PRO
Storage: 1 TB Seagate SSHD
Graphics: XFX Radeon HD 7870 GHz edition
PSU: XFX PRO550W

I'm in love with the Ryzen 7s for my next build
 
Solution
W
Sort by date Sort by votes
InvalidError

InvalidError

Titan
Moderator
May 18, 2007
23,815
5,265
108,890
The H60 may be worse than the 212 EVO. Did you try re-installing the 212 with new paste first? Even Intel's stock cooler should be able to keep the CPU under 80C at stock so if the 212 EVO is unable to keep yours any cooler, you have a major issue somewhere.
 
Upvote 0 Downvote
W

weberdarren97

Glorious
Aug 10, 2015
8,137
6
45,615


What is your voltage set at?



With how small the radiator is on the H60, it would likely be worse than the 212 EVO.



Ryzen 7 is more of a workstation lineup if you ask me. I would recommend Ryzen 5 for gaming.
 
Upvote 0 Downvote
P

papality

Honorable
Dec 1, 2012
74
0
10,640
Finally got my Arctic MX-4. Cleaned off the old stuff, applied the new paste. Stuck with the 212 and... first boot, temps stayed the same, Windows finally had enough and it started the resets. Got into the BIOS, made sure everything was set to minimums and defaults, windows wouldn't even load to desktop for more than 10,15 seconds. CPU Temp would be running for a few seconds and it would show multipliers up to 42x105, so could this be a motherboard issue in addition to CPU? Am I doomed?
 
Upvote 0 Downvote
W

weberdarren97

Glorious
Aug 10, 2015
8,137
6
45,615


Please set CPU base clock to 100MHz, not 105MHz.
 
Upvote 0 Downvote
P

papality

Honorable
Dec 1, 2012
74
0
10,640


It is set to 100 in BIOS, literally everything is at default settings. But when windows boots, that's what my tools are reporting, both coretemp and cpuid. And it waits about 10 seconds and resets

 
Upvote 0 Downvote
W

weberdarren97

Glorious
Aug 10, 2015
8,137
6
45,615




That is not default. Default is 34x100. Default turbo is 38x100. http://ark.intel.com/products/65520/Intel-Core-i5-3570K-Processor-6M-Cache-up-to-3_80-GHz

Let's try 36x100 for now. If this is stable, we can fine tune it later.
 
Upvote 0 Downvote
Solution
InvalidError

InvalidError

Titan
Moderator
May 18, 2007
23,815
5,265
108,890
If everything is back to default with the HSF correctly seated on an appropriate amount of correctly applied thermal paste and you are still seeing temperatures nearing 100C, then it could be a damaged CPU or just a bad temperature monitoring sensor. In HWInfo, how much package power does the VRM report? If the VRM says that power is normal and you still get high temperature, something clearly isn't right with heat transfer from the CPU to the HSF. Do all the cores read ~100C? If others read much lower, then it could be a fault only in that specific core.
 
Upvote 0 Downvote
P

papality

Honorable
Dec 1, 2012
74
0
10,640


Ok I have a stable boot at 36.0x99.9 per CPUZ, Coretemp, and HWInfo, temps polling in low to mid 40s idling.

In HWInfo, how much package power does the VRM report? If the VRM says that power is normal and you still get high temperature, something clearly isn't right with heat transfer from the CPU to the HSF. Do all the cores read ~100C? If others read much lower, then it could be a fault only in that specific core.
Click to expand...

I'm not entirely sure what this means? in HWI, the "CPU Package Power" row reads a minimum of 14.998 W, maximum of 26.512 W, and average of 15.93x.
 
Upvote 0 Downvote
InvalidError

InvalidError

Titan
Moderator
May 18, 2007
23,815
5,265
108,890

The only ways you could be hitting 100C while package power is only 16W would be if the HSF isn't making remotely decent contact with the CPU's IHS or if the temperature sensors themselves are messed up.

Triple-check that the HSF is truly installed correctly, flat on the CPU. If the HSF's base block still barely heats up, then a sensor issue is the most likely problem.
 
Upvote 0 Downvote
P

papality

Honorable
Dec 1, 2012
74
0
10,640


Sorry guess I wasn't clear, those numbers are from my current stable boot, multipliers set to 36x100 not what was a dying boot. from yesterday.
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom