F5F Stay Refreshed Hardware Desktop Why is my CPU running so hot?

Why is my CPU running so hot?

Why is my CPU running so hot?

F
fennecfox909
Junior Member
23
07-21-2016, 06:41 AM
#1
I bought a used i7 4770k as a budget upgrade from my previous i5 setup. I also got a Hyper 212 Evo cooler because it’s affordable and I’m familiar with them. After upgrading to Ran Heaven, Furmark, and RDR2, I didn’t notice much improvement in benchmarks. When I ran Prime 95 for a short time, the CPU cores were consistently reaching around 100°C. Could this be why there wasn’t a noticeable boost? I suspect the processor is throttling due to heat, trying to protect itself from damage. Since I’m new to PC building, I’m wondering if the cooling solution isn’t sufficient. My room temperature is usually between 72-74°F, so that seems unlikely. I also forgot to check the thermal paste before anyone asked, and my case has six fans (including an extra one I added). Could I have over-tightened the cooler? That’s a possibility I wasn’t sure about. Another thought is that the old thermal paste might need replacing. Just seeking advice from someone more experienced would be helpful.
F
fennecfox909
07-21-2016, 06:41 AM #1

I bought a used i7 4770k as a budget upgrade from my previous i5 setup. I also got a Hyper 212 Evo cooler because it’s affordable and I’m familiar with them. After upgrading to Ran Heaven, Furmark, and RDR2, I didn’t notice much improvement in benchmarks. When I ran Prime 95 for a short time, the CPU cores were consistently reaching around 100°C. Could this be why there wasn’t a noticeable boost? I suspect the processor is throttling due to heat, trying to protect itself from damage. Since I’m new to PC building, I’m wondering if the cooling solution isn’t sufficient. My room temperature is usually between 72-74°F, so that seems unlikely. I also forgot to check the thermal paste before anyone asked, and my case has six fans (including an extra one I added). Could I have over-tightened the cooler? That’s a possibility I wasn’t sure about. Another thought is that the old thermal paste might need replacing. Just seeking advice from someone more experienced would be helpful.

T
tylerkoch
Junior Member
26
08-10-2016, 06:07 PM
#2
Which thermal paste was applied? Which software do you use to track temperatures? Replace the cooler.
T
tylerkoch
08-10-2016, 06:07 PM #2

Which thermal paste was applied? Which software do you use to track temperatures? Replace the cooler.

Q
QwertyCat
Member
198
08-24-2016, 12:49 PM
#3
The Hyper 212 is a highly promoted cooler. It gets more credit than it deserves. Adding another fan would help move air through it. No, I didn’t notice you included a fan already. These 4770k chips are known for low-quality thermal paste under the IHS label, and replacing it usually makes a big difference—around 10-20°C lower temps possible. Skip the fake test results; this chip outperforms GPUs significantly. I’ve seen some cards fail with that software in the past.
Q
QwertyCat
08-24-2016, 12:49 PM #3

The Hyper 212 is a highly promoted cooler. It gets more credit than it deserves. Adding another fan would help move air through it. No, I didn’t notice you included a fan already. These 4770k chips are known for low-quality thermal paste under the IHS label, and replacing it usually makes a big difference—around 10-20°C lower temps possible. Skip the fake test results; this chip outperforms GPUs significantly. I’ve seen some cards fail with that software in the past.

I
iThinkitsPlum
Member
60
09-10-2016, 05:27 PM
#4
Reset the CMOS, but your BIOS might still assume an i5 is present. Test the connection by removing it and observing the thermal paste distribution.
I
iThinkitsPlum
09-10-2016, 05:27 PM #4

Reset the CMOS, but your BIOS might still assume an i5 is present. Test the connection by removing it and observing the thermal paste distribution.

M
mat_fram
Posting Freak
776
09-10-2016, 05:48 PM
#5
It could work well. I'll clear the CMOS and check the outcome.
M
mat_fram
09-10-2016, 05:48 PM #5

It could work well. I'll clear the CMOS and check the outcome.

N
NinjagobobMC
Member
69
10-01-2016, 01:42 AM
#6
I wasn't aware that Farkmouth had previously targeted GPUs. Finding a reliable delidding tool can be tricky, especially since some Amazon reviews mentioned severe damage to the CPU.
N
NinjagobobMC
10-01-2016, 01:42 AM #6

I wasn't aware that Farkmouth had previously targeted GPUs. Finding a reliable delidding tool can be tricky, especially since some Amazon reviews mentioned severe damage to the CPU.

S
StarSabre
Member
62
10-01-2016, 06:12 AM
#7
I'm checking HWMonitor for readings (maybe not perfect) and Arctic Silver 5 for compound temps.
S
StarSabre
10-01-2016, 06:12 AM #7

I'm checking HWMonitor for readings (maybe not perfect) and Arctic Silver 5 for compound temps.

F
Frinex10
Posting Freak
806
10-01-2016, 08:40 AM
#8
Because the plate is still being glued, your tool feels like a sharp edge. Make sure to watch tutorials and videos. You'll see which tools are recommended. Oh, FurMark is quite demanding. That's why it's called a stress test—people would overheat and damage their GPUs. I wouldn't recommend it at all. Instead, play your favorite AAA game and enjoy the experience in real time.
F
Frinex10
10-01-2016, 08:40 AM #8

Because the plate is still being glued, your tool feels like a sharp edge. Make sure to watch tutorials and videos. You'll see which tools are recommended. Oh, FurMark is quite demanding. That's why it's called a stress test—people would overheat and damage their GPUs. I wouldn't recommend it at all. Instead, play your favorite AAA game and enjoy the experience in real time.

_
_Sibex_
Junior Member
6
10-01-2016, 09:47 PM
#9
I thought using a specialized tool would increase my chances of success, even though I know some tools can cause mistakes too. I’m planning to research the method and try it out. It doesn’t look overly complicated based on what I’ve seen so far.
_
_Sibex_
10-01-2016, 09:47 PM #9

I thought using a specialized tool would increase my chances of success, even though I know some tools can cause mistakes too. I’m planning to research the method and try it out. It doesn’t look overly complicated based on what I’ve seen so far.

J
J1son
Member
211
10-01-2016, 11:14 PM
#10
Start with outdated CPUs before moving on. It's safer to test on older parts or those compatible with your board, allowing you to experiment without cost.
J
J1son
10-01-2016, 11:14 PM #10

Start with outdated CPUs before moving on. It's safer to test on older parts or those compatible with your board, allowing you to experiment without cost.