ChadEO

Distinguished
Jan 16, 2016
33
4
18,535
Please keep it to 1 thread.
Have a recently purchased 12400f CPU using the Intel stock cooler. I replaced the stock thermal paste with DeepCool Z5 and while the idle temps have improved and are now at 46-47, under load (playing Witcher 3 maxed out at 1080p) it still gets up to 99. It doesn't consistently hold 99 but spikes up to it. Clocks don't drop though but am just concerned about the temps. Case is a Cooler Master K350 with a Gentle Typhoon for rear exhaust, another Gentle Typhoon for bottom intake, and one of the older Corsair SP 120s on front intake. GPU is an OCPC RTX 3060ti that gets up to 84 under the same load. Are these normal temps or do I need to redo thermal paste application (used one pea in the middle, smaller peas on the corners)?
 
The differences between properly applied thermal pastes are minimal.

That's a 65 watt processor with 117 watts max turbo power, so says Intel. I wouldn't think it would be difficult to cool......but I don't have experience with the stock cooler.

I don't know anything about that case, but a quick Google says it has a lot of mesh, which should be conducive to air flow.

Haven't seen a mention of Gentle Typhoons in I guess 10 years. They were highly regarded at the time for being relatively quiet if I recall. Not sure if they move a lot of air?

Did all 3 of your fans come with the case?

Have you done any experimentation...moving the fans around; maybe turning one off? Or are you using a strictly stock configuration?

Have you fiddled with fan RPM settings in the BIOS? That would be my first idea.

Are you particularly noise-conscious?

Is there an over-riding reason you don't move to a better cooler? Or maybe even fans? Expense?
 
Use msi afterburner to show your power usage during gaming, the mobo probably pushes more power to the CPU than it needs.
If so then go into your bios and look through all the settings, disable any msi improvements that might be enabled and then make sure to set PL1 and PL2 to whatever max you think is ok for your cooling.

Second option your fan curve might be set to quiet, not cooling enough during high load.
 
  • Like
Reactions: ohio_buckeye
Definitely check to make sure the power limits are set as the poster above me says. If you get on YouTube there are usually lots of videos on how to do this step by step. Also you may consider turning the cpu fan to run 100%. I might be old school but many times that’s my preference if it doesn’t cause a lot of extra noise.
 

ChadEO

Distinguished
Jan 16, 2016
33
4
18,535
The differences between properly applied thermal pastes are minimal.

That's a 65 watt processor with 117 watts max turbo power, so says Intel. I wouldn't think it would be difficult to cool......but I don't have experience with the stock cooler.

I don't know anything about that case, but a quick Google says it has a lot of mesh, which should be conducive to air flow.

Haven't seen a mention of Gentle Typhoons in I guess 10 years. They were highly regarded at the time for being relatively quiet if I recall. Not sure if they move a lot of air?

Did all 3 of your fans come with the case?

Have you done any experimentation...moving the fans around; maybe turning one off? Or are you using a strictly stock configuration?

Have you fiddled with fan RPM settings in the BIOS? That would be my first idea.

Are you particularly noise-conscious?

Is there an over-riding reason you don't move to a better cooler? Or maybe even fans? Expense?
Nope, all three fans are aftermarket, case came with one Cooler Master branded intake. Haven't moved around the fans yet, this is just how I've had it set up for a while now. Yeah, tower is right beside me when in use so the noise of full blast fans is a no go. I already have plans to purchase a new cooler but still saving up for it now (still a broke student) so I'm just wondering if there's anything I can do in the mean time.
 

ChadEO

Distinguished
Jan 16, 2016
33
4
18,535
Use msi afterburner to show your power usage during gaming, the mobo probably pushes more power to the CPU than it needs.
If so then go into your bios and look through all the settings, disable any msi improvements that might be enabled and then make sure to set PL1 and PL2 to whatever max you think is ok for your cooling.

Second option your fan curve might be set to quiet, not cooling enough during high load.
Power draw in Afterburner maxes out at 59 so it shouldn't be that. Haven't thought about the CPU fan profile, I'll check that out
 

ChadEO

Distinguished
Jan 16, 2016
33
4
18,535
Definitely check to make sure the power limits are set as the poster above me says. If you get on YouTube there are usually lots of videos on how to do this step by step. Also you may consider turning the cpu fan to run 100%. I might be old school but many times that’s my preference if it doesn’t cause a lot of extra noise.
It gets really loud haha, even louder than the old Intel stock coolers but it seems to be worse in every other aspect except aesthetics
 
Nope, all three fans are aftermarket, case came with one Cooler Master branded intake. Haven't moved around the fans yet, this is just how I've had it set up for a while now. Yeah, tower is right beside me when in use so the noise of full blast fans is a no go. I already have plans to purchase a new cooler but still saving up for it now (still a broke student) so I'm just wondering if there's anything I can do in the mean time.

You can do quite well on a new cooler for $50 maximum if your case is wide enough to accept the better coolers.
 
I run a 12400F with the stock RM1 Intel cooler. Normally I use aftermarket coolers for noise/aesthetics, but when I did a rebuild it turned out the cutout in the case's motherboard tray didn't align with the new motherboard CPU location and the aftermarket cooler backplate fouled on the tray.

I installed the RM1 with the stock thermal pads on it, initially as a stop gap but it's been more than good enough so I've kept it. Doesn't get that loud in my system, maybe when driven hard but I game with headphones so don't notice. I might record fan-speed next time I play and see how high it gets.

The temperatures however never reach anything like yours; I've tested with Prime 95 Small FFTs and it levelled out at around 91°C after 10 - 15 minutes, and games don't push the CPU anywhere near as hard, as far as I know. Idle is more like 30°C I think, but then what's the ambient in your room?
 

ChadEO

Distinguished
Jan 16, 2016
33
4
18,535
I run a 12400F with the stock RM1 Intel cooler. Normally I use aftermarket coolers for noise/aesthetics, but when I did a rebuild it turned out the cutout in the case's motherboard tray didn't align with the new motherboard CPU location and the aftermarket cooler backplate fouled on the tray.

I installed the RM1 with the stock thermal pads on it, initially as a stop gap but it's been more than good enough so I've kept it. Doesn't get that loud in my system, maybe when driven hard but I game with headphones so don't notice. I might record fan-speed next time I play and see how high it gets.

The temperatures however never reach anything like yours; I've tested with Prime 95 Small FFTs and it levelled out at around 91°C after 10 - 15 minutes, and games don't push the CPU anywhere near as hard, as far as I know. Idle is more like 30°C I think, but then what's the ambient in your room?
I run an AC set at 22-24 degrees whenever I'm in here so it should be around that. I know the stock cooler is shit compared to aftermarket ones but I've never heard of 99 degrees being normal on a chip as low down the product stack as the 12400f which is why I'm confused. I might try to reseat the cooler since I changed motherboards and then repasted it with the DeepCool so I might just be getting the mounting wrong as I've always struggled with Intel's pins. But it doesn't seem to throttle at all which has me even more confused. Power draw remains the same, clocks remain the same, fps remains the same despite hitting 99.
 
I've never heard of 99 degrees being normal
The Intel OEM cooler was designed to be barely adequate. It is not unusual during some work loads for the CPU to reach 100°C and start thermal throttling. The performance drop can be very gradual. Most users that are not closely monitoring their CPU will barely notice the difference. If you want better temps you will need to buy a better cooler.

clocks remain the same
Many monitoring apps that sample the CPU MHz once per second do not accurately report the MHz when an Intel CPU first starts to thermal throttle. Thermal throttling can be changing the CPU speed hundreds of times per second.

Try running ThrottleStop and turn on the Log File option on the main screen. It will accurately record any thermal throttling and changes in MHz which other apps can miss. The TPL window will show what the turbo power limits have been set to by the BIOS. Some motherboards set the turbo power limits to Unlimited for maximum performance. This can create more heat than an Intel OEM cooler can manage.

Play a game for at least 15 minutes and then copy and paste the log file data to www.pastebin.com if you want me to have a look.

 

ChadEO

Distinguished
Jan 16, 2016
33
4
18,535
The Intel OEM cooler was designed to be barely adequate. It is not unusual during some work loads for the CPU to reach 100°C and start thermal throttling. The performance drop can be very gradual. Most users that are not closely monitoring their CPU will barely notice the difference. If you want better temps you will need to buy a better cooler.


Many monitoring apps that sample the CPU MHz once per second do not accurately report the MHz when an Intel CPU first starts to thermal throttle. Thermal throttling can be changing the CPU speed hundreds of times per second.

Try running ThrottleStop and turn on the Log File option on the main screen. It will accurately record any thermal throttling and changes in MHz which other apps can miss. The TPL window will show what the turbo power limits have been set to by the BIOS. Some motherboards set the turbo power limits to Unlimited for maximum performance. This can create more heat than an Intel OEM cooler can manage.

Play a game for at least 15 minutes and then copy and paste the log file data to www.pastebin.com if you want me to have a look.

Might try that tomorrow (it's 4am here) but I just did a 10 minute run of Prime95 and yep, it thermal throttled. Not hard, clocks just dropped from 3.9 to 3.5 at the lowest but it did. Oddly enough power draw has gone up to the mid 80s during the run. Still just using MSI Afterburner to monitor things. Prime 95 is still running and it hasn't caused my PC to crash at least.
 
Last edited:
Might try that tomorrow (it's 4am here) but I just did a 10 minute run of Prime95 and yep, it thermal throttled. Not hard, clocks just dropped from 3.9 to 3.5 at the lowest but it did. Oddly enough power draw has gone up to the mid 80s during the run. Still just using MSI Afterburner to monitor things. Prime 95 is still running and it hasn't caused my PC to crash at least.

Power draw in the mid 80W for a 65W CPU indicates a PL2 level usage.

MSI afterburner gives you time history graph. check for how long the CPU ran at mid 80s PL2 stage.

IIRC, the official intel spec is 8 sec and it is limited to 117W like said above.

You might want to find these settings and confirm in the BIOS:

PL1 - 65W

PL2 - 117W

Tau - 8 sec
 
Last edited:
The Intel OEM cooler was designed to be barely adequate. It is not unusual during some work loads for the CPU to reach 100°C and start thermal throttling.
That's traditionally been the view, but the more recent coolers have been a lot better:

In testing with a Core i5-12400, the cooler kept the CPU at a steady 73C with 20C ambient temperatures while running a full AIDA64 stress test, resulting in around 80W of power consumption. According to the outlet, the only downside was the fan's rather high 3100 RPM, which produced a lot of wind noise.
https://www.tomshardware.com/news/new-alder-lake-stock-heatsink-tested

For example, the Laminar RM1 is more efficient (at the same noise level it has higher cooling performance) than the Arctic Alpine 17, and the Laminar RS1, a class below it, is not much worse.

That second link, they test the i5 12400 with the RS1 stock cooler (which is poorer than the RM1 the i5 actually comes with) and it held at 82°C in Cinebench R23 at max RPM.

As I've said, I can't reach 100°C in mine when running Prime95 small FFT. Ambient is 18 - 20°C. There are certainly better coolers out there, but it's old thinking that the stock coolers are no good at high workloads. What the OP's seeing isn't right.
 
  • Like
Reactions: uWebb429

ChadEO

Distinguished
Jan 16, 2016
33
4
18,535
Power draw in the mid 80W for a 65W CPU indicates a PL2 level usage.

MSI afterburner gives you time history graph. check for how long the CPU ran at mid 80s PL2 stage.

IIRC, the official intel spec is 8 sec and it is limited to 117W like said above.

You might want to find these settings and confirm in the BIOS:

PL1 - 65W

PL2 - 117W

Tau - 8 sec
Will try to look for it again but I'm still trying to figure out the MSI bios, been used to Asus ever since I got my first pc.
 

ChadEO

Distinguished
Jan 16, 2016
33
4
18,535
That's traditionally been the view, but the more recent coolers have been a lot better:


https://www.tomshardware.com/news/new-alder-lake-stock-heatsink-tested



That second link, they test the i5 12400 with the RS1 stock cooler (which is poorer than the RM1 the i5 actually comes with) and it held at 82°C in Cinebench R23 at max RPM.

As I've said, I can't reach 100°C in mine when running Prime95 small FFT. Ambient is 18 - 20°C. There are certainly better coolers out there, but it's old thinking that the stock coolers are no good at high workloads. What the OP's seeing isn't right.
Is it possible my CPU or motherboard (not sure which one it comes with) shipped with faulty sensors? If so, would it be worth getting the warranty or will it not affect performance at all
 
Is it possible my CPU or motherboard (not sure which one it comes with) shipped with faulty sensors? If so, would it be worth getting the warranty or will it not affect performance at all
It's possible, but I'd say unlikely. Each core has its own sensor so they'd all have to be wrong (or interrogated wrong). If either was faulty I would definitely replace it.

I selflessly played Horizon Zero Dawn last night for a couple of hours while logging everything, which based on age/recommended specs should be more demanding on the CPU than Witcher 3. 1080p / max settings. CPU core temps were 60 - 70°C with the very occasional spike to 72°C. No throttling ever, mostly operating at 4 GHz across all cores. CPU fan speed was around 1750 rpm, occasionally going over 2000 rpm.

Are you sure the cooler is seated correctly? I vaguely remember a thread here where it turned out one of the corner push-pins hadn't quite been pushed fully into place - I remember installing mine, the last one seemed to require more force than is comfortable to completely click home, and just before it did it felt like it was already in, if you see what I mean.



Edit: this may have been the thread. The OP could see the pins retract a little when pressing on the opposite corner, which was a sure sign that they weren't fully home.
 
Last edited:

ChadEO

Distinguished
Jan 16, 2016
33
4
18,535
It's possible, but I'd say unlikely. Each core has its own sensor so they'd all have to be wrong (or interrogated wrong). If either was faulty I would definitely replace it.

I selflessly played Horizon Zero Dawn last night for a couple of hours while logging everything, which based on age/recommended specs should be more demanding on the CPU than Witcher 3. 1080p / max settings. CPU core temps were 60 - 70°C with the very occasional spike to 72°C. No throttling ever, mostly operating at 4 GHz across all cores. CPU fan speed was around 1750 rpm, occasionally going over 2000 rpm.

Are you sure the cooler is seated correctly? I vaguely remember a thread here where it turned out one of the corner push-pins hadn't quite been pushed fully into place - I remember installing mine, the last one seemed to require more force than is comfortable to completely click home, and just before it did it felt like it was already in, if you see what I mean.



Edit: this may have been the thread. The OP could see the pins retract a little when pressing on the opposite corner, which was a sure sign that they weren't fully home.
I've tried to reseat the cooler more than once and I'm like 90% sure it's properly seated now since I don't see any movement when I push on the pins. I just went ahead and ordered a Peerless Assassin Mini anyway just to save myself anymore headaches because of this haha
 
  • Like
Reactions: Lucky_SLS

ChadEO

Distinguished
Jan 16, 2016
33
4
18,535
Got a Thermalright PA120 Mini and ran Prime95 for 2 hours, peaked at 92 degrees which is still quite high. At this point it's either my case has just terrible airflow or the CPU isn't making good contact with whatever cooler I put on it
 

ChadEO

Distinguished
Jan 16, 2016
33
4
18,535
Build:

CPU: 12400f

Cooler: Thermalright Peerless Assassin Mini 120
Thermal Paste: Thermal Grizzly Kryonaut

Motherboard: MSI Pro B760-P Wifi DDR4

Case: Cooler Master K350

Case Fans: Front intake, bottom intake, rear exhaust. All case fans are Arctic P12s



Idle temps are in the 30s, high 30s if I had already pushed the CPU since boot, low 30s if not. Ran Prime95 for 2 hours and temps held stable at 92 degrees with 0 throttling. With the stock cooler I was reaching 100 and throttling but not enough to cause crashes, just lowered clock speeds. Idle on the stock cooler was in the 40s. Aren't these temps still high considering the CPU?