Question i9-9900x max temps

[shep66]

Hi,

I have what I think is a simple question about the maximum safe temp of this CPU. Due to on going problems with PC crashes during gaming, basically when it's under heavy load, I've done a lot of troubleshooting and am at the point where I'm trying to rule out an over heating CPU as the cause. (The BSOD in question is DPC_WATCHDOG_VIOLATION, which is apparently an issue ported over from older versions of Windows and apparently stems from an outdated SSD driver. I've updated both the driver and firmware. However, just as often there is no BSOD at all, just an unprompted PC restart.)

I've used two different monitoring tools and both show various cores regularly hitting 94C. I know the core temp is usually higher than the package temp but this is still in excess of the advertised maximum safe operating temp of 92C. Is 92C really the ultimate red line? And by default the CoreTemp UI displays TjMax = 110C. Que?

All I've read online suggests that, even under load, it shouldn't be going above 80C on a regular basis. That sounds right to me, yet the firm I bought the OCed bundle from say:
"As for the temperatures you are reporting under load, they certainly aren't low, but they are not high enough to cause any issues, at 95C and above the CPU will throttle performance, but it should not cause blue screens."
Now, I love this company but is that really true?

 

[Dark Lord of Tech]

How To - Intel CPU Temperature Guide

Update: March 12th, 2024 Preface The topic of processor temperatures can be very confusing. Conflicting opinions based on misconceptions concerning terminology, specifications and testing leaves users uncertain of how to properly check cooling performance. This Guide provides an understanding...

forums.tomshardware.com

Great read here , one of the mods maintains it.

 

[CompuTronix]

shep66,

Maximum Junction Temperature (Tj Max) or "Throttle" temperature for the i9-9900X is in fact 92°C:

Intel Product Specifications - https://ark.intel.com/content/www/u...es-processor-19-25m-cache-up-to-4-50-ghz.html

As the 9900X is a 10 Core / 20 Thread 165 Watt TDP processor, and as your H115i Pro is a 280mm AIO, it's somewhat less than adequate to cool your CPU with all Cores at 4.4, especially with a conservative overclock. Even the 360mm H150i Pro would be just marginal. HEDT (High End DeskTop) processors such as yours really need very high end cooling; preferably a custom loop.

While it's entirely possible that your BSOD's may be coincidental with issues other than excessively high Core temperatures, more information is needed:

(1) Has this problem appeared relatively suddenly, or has it gradually become more frequent over time?

(2) How old is your rig, as well as the H115i Pro?

(3) Which Power Supply is installed?

(4) What is your ambient (room) temperature? ("Standard" or normal is 22°C or 72°F.)

(5) What are your idle temperatures, and is "idle" really idle at only 1 or 2% CPU Utilization in Windows Task Manager?

(6) Which versions of what monitoring utilities are you using? ("Package" temperature typically corresponds with the hottest "Core".)

(7) Other than gaming, have you run any stress tests?

(8) Have you tried running a fresh install of Windows on a spare drive to rule out software?

Although most processors Throttle at 100°C or 212°F, it’s not advisable to run your CPU near it's thermal limit. If your hottest Core is near it's specified Tj Max Throttle temperature, then your CPU is already too hot. The consensus among well informed and highly experienced system builders, reviewers and overclockers, is that cooler is better for ultimate stability, performance and longevity.

Experts all agree that it's prudent to observe a reasonable thermal margin below Tj Max, so here's the nominal operating range for Core temperature:

Core temperatures above 85°C are not recommended.

Core temperatures below 80°C are ideal.



Keep in mind that Core temperatures increase and decrease with Ambient temperature.

CT 😎

 

[shep66]

Thank you both for the comprehensive reply.

(1) Has this problem appeared relatively suddenly, or has it gradually become more frequent over time? The restarts appeared relatively suddenly with the previous mobo bundle (see below) and the BSODs appeared relatively suddenly on the current build.

(2) How old is your rig, as well as the H115i Pro? - The mobo bundle, including the cooler, is brand new. I had random restarting issues (but no BSOD) on the previous bundle (had a 20% OC on a i7-5930k and H100 cooler), which I put down to the graphics card (since RMAed). This seemed logical, since I had no issues at all with 2 x GTX 980ti - it was the introdction of the new RTX 2080ti that coincided with these random restarts during firefights and other massive conflagrations in BFV. The random restarts are still occurring but the BSOD is new. The common theme is that both only ever happen under load. The only PC component common to both builds, apart from the case, is the PSU. I have bought a spare to swap out but not unboxed it yet (depends on exhausting other avenues first). But the fact remains that the temps are too high, except Scan has just replied to my enquiry to say that 95C is normal. Who is right? And are they matching the wrong coolers with both these CPUs?

(3) Which Power Supply is installed? - Corsair HXi 1200W (was running SLI, probably overpowered even then)

(4) What is your ambient (room) temperature? ("Standard" or normal is 22°C or 72°F.) - 20-22C

(5) What are your idle temperatures, and is "idle" really idle at only 1 or 2% CPU Utilization in Windows Task Manager? - Idle at 35C and all cores between 0-2% load.

(6) Which versions of what monitoring utilities are you using? ("Package" temperature typically corresponds with the hottest "Core".) - CoreTemp 1.14

(7) Other than gaming, have you run any stress tests? - tried running Prime95 but PC locked up as soon as I clicked the Run/Start button. Scan said Prime95 isn't suited to Intel CPUs (they use Aida64), which I found an odd thing to say, since I've run it on previous Intel CPUs with normal results.

(8) Have you tried running a fresh install of Windows on a spare drive to rule out software? - Yes, happened on a previous SSD and have also re-installed Win10 on this SSD.

 

[shep66]

I think I'll move this thread elsewhere as it doesn't look like the hot CPU is the root cause. Following some more troubleshooting I'm now even less certain I understand what's going on with this PC.

For the record, I swapped in a new Corsair RM750X PSU but instead of bolting it inside the case I left it sitting out, resting on the outside of the case that I laid flat on the desk while I thrashed BFV for several hours. Problem solved. Until I installed it inside the case, closed the side panel, stood the case upright and then the BSODs started again. So I unscrewed it from inside the case and just put it back where I'd had it before, case lying on its side with the side panel off once again, spent over two hours on BFV and no BSODs.

For good measure I swapped the old PSU back in, also leaving it outside of the case but it kept on with the BSODs. I was starting to think the case might be a factor but now I'm clueless.

Throughout all this the CPU temps are still high, up to 95C regardless of the PSU configuration, so this doesn't seem to be a factor in whether or not the BSODs happen.

The last thing I'll try with the PSU is swapping out all the old cables for new ones, something I should've done anyway (I have already installed the new 24-pin & 12-volt cables but not the new PCIe and peripheral/SATA cables).

 

[CompuTronix]

Intermittent problems can indeed be the most difficult and time consuming to troubleshoot, especially on new parts which have no established history of reliability, such as known-good used parts. Based upon your descriptions, it seems that the BSOD's are somehow related to the graphics card and physical position of the motherboard relative to horizontal versus vertical.

Since you're using an AIO, the waterblock / pump unit doesn't impose significant cantilever forces on the CPU socket like a heavy air cooler such as the Noctua NH D15 at nearly 3 pounds. Nonetheless, that doesn't rule out the potential effects of cantilever forces imposed on the PCIE X16 slot by your RTX 2080ti, or on the card itself.

If you still have your GTX 980ti's, have you tried running one of those cards to rule out a defective RTX 2080ti XC?

I suggest that you return to the previous configuration where you had the case on it's side using the replacement PSU externally, then see if you can duplicate your successful 2 hour gaming run. If it again doesn't BSOD, then reassemble the case in the original configuration with your replacement PSU installed.

With the case positioned on it's side, perform another gaming run. If it still doesn't BSOD, then stand it up vertically and retest it. If it BSOD'd in the vertical position, it would seem to point toward a damaged / defective PCIE X16 slot.

Lastly, move the graphics card to the second PCIE slot and test your rig once again in both the horizontal and vertical positions. If it no longer BSOD's in either position, then the problem has to be related to the motherboard's PCIE X16 slot.

CT 😎