HP OMEN 45L 4090 and 14900k completely crashes when playing certain games for brief period

[_Keven]

Hello,

I've been having issues with this PC recently where when playing certain games for some time 20+ minutes the entire pc turns off almost as if it lost power. Sometimes it takes several minutes before I can successfully turn it on, and sometimes it reboots itself after a few seconds. Windows event viewer doesn't show anything about the crash (likely because my entire pc crashed including event viewer). 

I ran HWinfo64 and logged all of the hardware data leading up to a crash but its a little overwhelming and hasn't helped me diagnose the problem. CPU temps before the crash seemed normal (around 65-70C) and so did the GPU temps (65C, GPU Hot spot was at 76C). I think its a power issue but I have no idea how to test for that. Any help would be greatly appreciated. I've had this PC for 9 months and wasn't expecting to have to replace it.

Note: I recently upgraded my monitor to a 1440 OLED 480 HZ. I think this might've gotten worse since I upgraded? I'm not sure exactly. I also have 2 other monitors plugged in but are both standard specs. 

[Bill_To]

Greetings @_Keven 

Welcome back to the Forum.

Your PC's symptoms appear to be a hardware problem.

CPU and GPU Temps per HWiNFO look very good.  Have you also checked voltages and : system board, VRM, memory, and M.2 temps?

Nvidia has been having some driver issues lately with the latest drivers. You might want to try an older Nvidia driver.

W11 has a "non-destructive" in-place upgrade option using Windows Update if you are interested in trying this. This option is in Settings>System>Recovery>Fix problems using Windows Update. Back up data on the system drive before doing any OS modifications.

Have you tried running HP startup diagnostics?

Check your PC's hardware using HP Diagnostics as follows:

Start the PC. Repeatedly tap the "ESC" key.

Select "F2". Run extensive system and component tests to confirm: a drive problem, a memory problem, or other component problem.

Your PC has a 14900K. You might want to check this Site for tips on testing the CPUs condition. Try all troubleshooting tips at this site.

I can't see using three monitors with a 4090 as a problem but try one monitor to see what happens.

You would have to try a different power supply if you cannot ID the cause of the problem by trying the above troubleshooting tips.

And you should have the HP warranty option if your nine month old PC is in warranty.

Regards

[_Keven]

Hi Bill,

After doing further testing and logging using HWinfo I noticed I was experiencing thermal throttling 12 times during a 30 minute session of gaming. This seemed alarming so I did more benchmark tests to see if my CPU's performance was being impacted. I ran Cinebench r23 and got a comically low score of 25285. Others with the same processor are reporting scores in the 40,000s. 

I'm no expert but from what I've read so far slightly undervolting might give my CPU a more consistent performance and thus reduce thermal throttling. But I can't seem to figure out how to undervolt the CPU. Using intel's extreme tuning tool it only lets me increase my core voltage offset, rather than decreasing it. The Omen Gaming hub has the same issue. I tried going into BIOS and looking for some sort of undervolt lock but couldn't find anything that allowed me to undervolt the CPU. Is there anything else I can do to try and undervolt it?

[Bill_To]

Greetings @_Keven 

The CPU is thermal throttling.

HP generally do not provide BIOS CPU tweaks. And I really can't see HP applying CPU voltages above Intel's recommendations at the factory.

The only way to get around this roadblock, I am aware of, would be to get a retail Intel "Z" MB which will provide granular CPU tweaking options.

Did the Intel Processor Diagnostics Test pass?

Have you checked the existing CPU TIM application? Maybe repasting the CPU might help with thermals.

Many folks recommend a 360 mm AIO CPU liquid cooler for this processor. What size radiator does your PC have?

Regards

[_Keven]

Hey Bill,

That's really unfortunate about the lack of CPU tweaking available. Later today I guess I'll open it up and check the thermal paste and reapply some. I'll let you know the radiator dimensions when I do that as well. Thanks for the quick response though!

[_Keven]

Hey Bill,

I thought this might be interesting/helpful for you. I was going to apply new thermal paste onto the CPU, but before doing so I wanted to do some benchmark tests using Cinebench R23. That way I could compare the results with how it performs after applying the paste.

I did one test with a balanced fan setting mode (automatic) and CPU temps got up to 97 at the peak. Very high I know. I ran another test on turbo mode (automatic fan settings but just faster I guess) and temps reached a max of 85. The benchmark score was also significantly better. I then did a final test with fans near max speed (>80%). About halfway through the minute or so long test my PC crashed again in the same fashion as before. While this happened I was reading the HWinfo CPU temps and they were relatively low around 80 C. I decided to try the test again and this time it completed without any crashes and the max CPU temp was 81.  Note this was also by far the best benchmarket score of 32349 (which I've heard is still very low for this processor). 

I'm surprised my PC crashed when the fans were going the fastest, as the CPU temp was the lowest on average throughout the test. So I don't think this is a CPU heating problem but I'm not really sure where else to go from here. Could it be a bad PSU? GPU is faulty? idk. What do you think of this info?

[Bill_To]

Greetings @_Keven 

Thanks for the additional info. Good troubleshooting steps.

A finely tuned 14900K can hit around 42000 or higher when running Cinebench R23. I would not get hung up on benchmarks. It takes a lot of BIOS tweaking, which you cannot do with the HP MB. You want a stable PC which does not crash.

It's tough to diagnose the type of PC crashing issue you are having. A faulty power supply could cause the problem (possible CPU and GPU crashing) you are experiencing.

You could also primarily check the GPU by running the FurMark stress test to see if the 4090 has a problem.

R23 primarily stresses the CPU. I can see high fan speeds allowing the CPU to hit higher clock speeds thus drawing more power. So you might have an over voltage problem.

Did you run the Intel Processor Diagnostic Tool I referenced in a prior response? This test will let you know if the 14900K has a problem.

I build PCs and always check component stability using OCCT free. This software might help to figure out what is happening.

Regards

[_Keven]

Hey Bill thanks for the response,

I'll do those tests first thing when I get home. I have run furmarks in the past and I did experience the crash after 5-6 minutes. But the fact that this is crashing during Cinebench as well leads me to believe its a power supply issue maybe (as cinebench is cpu intensive, and furmarks is gpu intensive if I have that right)? I'll keep you posted.

[Bill_To]

Greetings @_Keven 

My pleasure.

You are correct concerning FurMark and Cinebench.

Make sure you do the Intel test to confirm the 14900K is okay.

Regards

[_Keven]

So I ran the Intel diagnostic tool and it passed all the tests. I can reliably cause the crash running furmarks at 1440 in roughly 7-10 minutes. So I'm guessing its a gpu problem, but its still odd I had that crash on cinebench that one time. Anyways I'm going to take it into microcenter this Friday, and have them fix it. Its still under warranty, but if they do in fact have to replace the 4090 its probably going to be a while since they're seemingly always out of stock. Thanks for all the help though! I've learned a lot throughout all of this.