[SOLVED] Strange i5-9400f temps

Feb 6, 2020
3
0
10
I have a really strange situation and I'm not sure what's going on, so I wanted to make sure everything is OK.

I have:

i5-9400f
EVGA 120 CLC AIO cooler
Cooler Master Q300L case

The system idles at 25 - 30 C. Running Cinebench R20, I can hit 3.9 - 4.0 GHz on all cores and it never reaches 50 C on temps.

I thought this was strange, so I switched the CLC out with the stock fan. Temps hit 80 - 85 C with very little effort.

So my question: Is this normal? Especially with this case? I've heard bad things about air flow with it, but so far it rocks. Is it possible something isn't reading correctly?


Thanks in advance!
 
Solution
... i5-9400f ... EVGA 120 CLC AIO cooler ... Cinebench R20 ... never reaches 50 C ... switched the CLC out with the stock fan. Temps hit 80 - 85 C ... Is this normal? ... Is it possible something isn't reading correctly?

... One core hit 49 C in AIDA64 ... I would say there's a sensor mucked up, but the air cooler ran much higher readings.
Foofah,

Intel's stock coolers are indeed very poor. What you're observing is completely normal and expected.

Intel's specification for Digital Thermal Sensors (DTS or "Core temperature sensors") accuracy is 5°C across the entire operating range. The sensors are typically very reliable.

If CineBench R20 (File > Preferences > Minimum Test Duration) is set for 600 seconds...

Barty1884

Retired Moderator
Cinebench is a relatively short term, and an AIO takes time to reach equilibrium (liquid heats up over time).

You'd want to run a long-term test to determine 'true' temperatures.... something like Prime95* or Aida64* for at least 30 minutes

Even gaming for an hour would give you a better idea of 'true', real world temperatures.

While a 9400F isn't particularly hot, at stock speeds..... I'd still expect to see a little higher temps (~60'C) on a 120mm AIO.

*Both Prime & Aida include AVX instruction sets, which will push your temps higher than you're ever likely to see in 'real world' use, unless you perform AVX workloads.
 
Feb 6, 2020
3
0
10
Cinebench is a relatively short term, and an AIO takes time to reach equilibrium (liquid heats up over time).

You'd want to run a long-term test to determine 'true' temperatures.... something like Prime95* or Aida64* for at least 30 minutes

Even gaming for an hour would give you a better idea of 'true', real world temperatures.

While a 9400F isn't particularly hot, at stock speeds..... I'd still expect to see a little higher temps (~60'C) on a 120mm AIO.

*Both Prime & Aida include AVX instruction sets, which will push your temps higher than you're ever likely to see in 'real world' use, unless you perform AVX workloads.

Thanks for the reply! I'll give it a harder workout later and see what happens.
 
Feb 6, 2020
3
0
10
Ran AIDA64 for over an hour, then played Starcraft II and FFXIV for over three hours total after that. One core hit 49 C in AIDA64 for a moment, but that's as high as I could get it to go.

I would say there's a sensor mucked up, but the air cooler ran much higher readings.
 

CompuTronix

Intel Master
Moderator
... i5-9400f ... EVGA 120 CLC AIO cooler ... Cinebench R20 ... never reaches 50 C ... switched the CLC out with the stock fan. Temps hit 80 - 85 C ... Is this normal? ... Is it possible something isn't reading correctly?

... One core hit 49 C in AIDA64 ... I would say there's a sensor mucked up, but the air cooler ran much higher readings.
Foofah,

Intel's stock coolers are indeed very poor. What you're observing is completely normal and expected.

Intel's specification for Digital Thermal Sensors (DTS or "Core temperature sensors") accuracy is 5°C across the entire operating range. The sensors are typically very reliable.

If CineBench R20 (File > Preferences > Minimum Test Duration) is set for 600 seconds (10 minutes), it'll run uninterrupted at a fairly steady workload close to 100%, which allows thermals to stabilize.

AIDA64 has 4 CPU related stress test selections (CPU, FPU, Cache, RAM) which have 15 possible combinations that yield 15 different Core temperatures, so you need to be very specific concerning which test(s) you actually ran. If it was just the individual CPU test, it's only about 70% workload.

Prime95 v29.8 Small FFT's with all AVX selection disabled is a true steady-state 100% workload. Prime95 with AVX is a 130% workload.

“Stress” tests vary widely and can be characterized into two categories; stability tests which are fluctuating workloads, and thermal tests which are steady workloads. Prime95 v29.8 Small FFT's (AVX disabled) is ideally suited for testing thermal performance, because it conforms to Intel's Datasheets as a steady 100% workload with steady Core temperatures. No other non-proprietary utility can so closely replicate Intel's thermal test workload.

n7sRDt4.jpg


Note: Click on the AVX test selections that are not greyed out so that all three AVX boxes are checked, as shown above.

Utilities that don't overload or underload your processor will give you a valid thermal baseline. Here’s a comparison of utilities grouped as thermal and stability tests according to % of TDP, averaged across six processor Generations at stock settings rounded to the nearest 5%:

Ydre0YW.jpg


Although these tests range from 70% to 130% TDP workload, Windows Task Manager interprets every test as 100% CPU Utilization, which is processor resource activity, not actual workload. Core temperatures respond directly to power consumption (watts), which is driven by workload. Prime95 v29.8 Small FFT’s (AVX disabled) provides a steady 100% workload. If Core temperatures don't exceed 80°C, your CPU should run the most demanding real-world workloads without overheating.

CT :sol:
 
Solution