2.71
#...
·
·
2
likes
|
|---|
|
Not sure which forum this goes in but… So I have a newish pc. i7 13700f with a 4070ti. Less than 2 years old. It’s a locally bought pre-built based off a deepcool case/aio and a gigabyte mb. Just for kicks, I thought I would run a hardware monitoring app while pixinsight was running. My CPU would hit 100 degrees C within SECONDS and start to thermal throttle!!! (on a side note, I found out my ram was underclocked as well lol) Turns out, CPU thermal paste has a lifespan, and that lifespan is shortened by running your CPU at 100%. I got some decent (and cheap) thermal paste of amazon, and now the CPU usually hangs between 75 and 85C. No throttling. Might be worth checking up on your temps. |
9.13
#...
·
·
1
like
|
|---|
| Thermal throttling is kinda normal with the new Intel processors. (It doesn't hurt them) They're pushing upwards of 300 watts of power. I have a 360mm rad and still get thermal throttle during long stacking sessions. I wouldn't be surprised if your thermal paste was fine, but your cooling solution was inadequate. |
|
2.71
#...
·
|
|---|
| Everything is the same except the paste ¯\_(ツ)_/¯ |
|
9.13
#...
·
|
|---|
Sean Mc: May I ask what you're using for cooling? Just curious. |
3.15
#...
·
|
|---|
| Your thermal paste may have been incorrectly applied by the company that built your PC. I have had the same thermal paste for 6 years now, and never seen thermal issues like this, it really shouldnt go bad on its own. |
|
2.71
#...
·
|
|---|
|
I’m not sure exactly which model my cooler is, but it’s a deepcool, 240 or 280. The cpu is a 13700f so 220w max. the thermal paste is pre-applied to the cooler so it’s unlikely that the builder screwed it up. most of what I’ve read states that thermal paste dries out over time and loses it’s effectiveness. Running max loads accelerates the degradation. Mist articles state that thermal paste lifespan averages 2-3 years ¯\_(ツ)_/¯ |
|
9.13
#...
·
|
|---|
|
I've been building PCs for decades and I've never had to replace thermal paste. I even did crypto mining and rebuilt dozens of graphics cards. Under those constant huge loads the thermal paste was never the failure point ever (thermal pads were). This held true even for the old Vega SOC cards which have memory and cpu on the same heatsink, those things can really pump the heat! A 240 or 280mm radiator is likely too small for that thermal load. I have a 360mm on my 12900k and it still eventually throttles. I went with the 360mm right off the bat because of alot of reviews/warnings about the wattage of these processors. Intel really turned up the juice to compete with AMD instead of releasing new dies. Researching about your 13700 says it has a higher rated max TDP than my 12900. If I were to do it all again I would have gone with an open loop system for increased thermal capacity. There's a small chance that you had this issue from the start, since you said you only recently started monitoring. Uneven pressure on the die can cause this, or like you said, it could have just been bad paste (from the beginning) . You probably just never put a big enough load on the processor to notice prior. The reality is Intel just runs hot as hell, and I wished I had went AMD instead. |
2.15
#...
·
·
1
like
|
|---|
| Just as an anecdote, I recently finished building my new PC with an i7-13700KF and a 4080 Super. I use a twin tower air cooler and only see thermal throttling under extreme stress tests (I did not OC btw, just running at stock speed, slightly undervolted). Even when thermal throttling, my CPU would only dip to about 4.7-5 GHz, so I would definetly check that the thermal paste is correctly applied or that something stupid wasn't done, such as not removing the plastic cover over the thermal block. Just my two cents. I would also agree that these chips just draw much more power compared to older cpus. My old machine only had a single 8 pin CPU socket on the MOBO while my current one has two. |
18.84
#...
·
|
|---|
|
There is no need to replace the paste due to aging. I use an i9-13900KS and it runs into throttling after seconds in PI WBPP, though I use an external water cooling tower with 3 fans which easily can radiate more than 300W. The water temperature difference is less than 4°C. The problem is the too small CPU heat spreader size, which cannot not transport the heat to the cooler. The energy density is too high. Additionally the Intel CPU have a concave cap, and the thermal transfer is not optimal. You have to use a convex cooler. AMD is reversed. It is good idea to asses the paste imprint to verify the are of contact. Long story short: Don't worry it is normal. Of course if the paste is applied too thick, or too thin, then of course reapplying it, my help - but not for aging. Example imprint:  Shaped cooler:  |
