Question Intel Turbo Boost causes BSOD crashes with an i9-13900K CPU ?

[sketcheyes]

So, as of recently (around 1 month) I have been getting many, and I mean many BSODs. I started to look into potential causes and I am sure it is the Intel Turbo Boost option. The BSODs would not typically occur in game though which confuses me. It would be when I am just browsing or even if I am idle. The most often error code I get is KMODE_EXCEPTION_NOT_HANDLED. Things I have tried already are:

• Resetting the PC
• Using Intel XTU and turning down the Core Multiplier
• Resetting BIOS
• Running MemTest (no errors were found)

Specs:
https://prnt.sc/_ct5RtxcEdMN

Full PC Page:

https://www.microcenter.com/product/662597/powerspec-g472-gaming-pc

 

[ubuysa]

There is a well known flaw with high end 13th and 14 gen Intel CPUs that causes the CPU to ask for higher voltages than it can handle. If you do nothing you will cook your CPU. Running a turbo boost is very likely to cook the CPU even faster. Once cooked the CPU is scrap. Intel, or your PC vendor, will replace it under warranty, but that will take time.

Intel have released two microcode updates that stop 13th and 14th gen CPUs from seeking too high a voltage and these two updates prevent the problem happening, but they will not fix an already cooked CPU. These microcode updates are shipped as BIOS updates, look for microcode updates x125 and x129 in the BIOS update description. You need to update to the x125/x129 BIOS now.

 

[sketcheyes]

There is a well known flaw with high end 13th and 14 gen Intel CPUs that causes the CPU to ask for higher voltages than it can handle. If you do nothing you will cook your CPU. Running a turbo boost is very likely to cook the CPU even faster. Once cooked the CPU is scrap. Intel, or your PC vendor, will replace it under warranty, but that will take time.

Intel have released two microcode updates that stop 13th and 14th gen CPUs from seeking too high a voltage and these two updates prevent the problem happening, but they will not fix an already cooked CPU. These microcode updates are shipped as BIOS updates, look for microcode updates x125 and x129 in the BIOS update description. You need to update to the x125/x129 BIOS now.

Yes, I have already updated to the latest version. Same problem persists with Turbo Boost Enabled. Turned it off and no problems whatsoever.. other than the performance drawbacks

 

[ubuysa]

If it only BSODs with turbo boost on then I'd be pretty confident it was a CPU issue. Let's get some real data though...

Can you please download and run the SysnativeBSODCollectionApp and upload the resulting zip file to a cloud service with a link to it here. The SysnativeBSODCollectionApp collects all the troubleshooting data we're likely to need. It DOES NOT collect any personally identifying data. It's used by several highly respected Windows help forums (including this one). I'm a senior BSOD analyst on the Sysnative forum where this tool came from, so I know it to be safe.

You can of course look at what's in the zip file before you upload it, most of the files are txt files. Please don't change or delete anything though. If you want a description of what each file contains you'll find that here.

 

[sketcheyes]

Yes, here is a link of the zip file:

https://www.mediafire.com/file/ayuukqeicdv0anp/SysnativeFileCollectionApp.zip/file

 

[ubuysa]

Your RAM is overclocked at 5600MT/s, and although that's within both the RAM capability and the CPU capability I'd like you to remove that overclock and run the RAM at the SPD speed of 4800MT/s. This is so that we can be sure that RAM is not an issue.

I'm asking this because taken together all the dumps could be RAM related and at CPU turbo speeds flaky RAM may be more likely to fail.

Although you've tested your RAM, no memory tester can be 100% accurate, so if it still BSOD with the RAM at 4800MT/s then remove one RAM stick for a few days, or until you get a BSOD. Then swap sticks and run on the other stick for a few days, or until you get a BSOD. Be sure to consult your motherboard manual to ensure that the one stick is in the correct slot.

If it still BSODs at 4800MT/s and on both sticks running on their own we'll consider your RAM to be good and move on.

 

[sketcheyes]

Your RAM is overclocked at 5600MT/s, and although that's within both the RAM capability and the CPU capability I'd like you to remove that overclock and run the RAM at the SPD speed of 4800MT/s. This is so that we can be sure that RAM is not an issue.

I'm asking this because taken together all the dumps could be RAM related and at CPU turbo speeds flaky RAM may be more likely to fail.

Although you've tested your RAM, no memory tester can be 100% accurate, so if it still BSOD with the RAM at 4800MT/s then remove one RAM stick for a few days, or until you get a BSOD. Then swap sticks and run on the other stick for a few days, or until you get a BSOD. Be sure to consult your motherboard manual to ensure that the one stick is in the correct slot.

If it still BSODs at 4800MT/s and on both sticks running on their own we'll consider your RAM to be good and move on.

Update: Set ram to 4800 with turbo boost on and still blue screened. Took out one ram stick and tested single ram sticks at a time and still experience same issue with both. Just takes longer to blue screen on average with single than with both on. However, windows is much more buggy. Sfc scans don’t work. Webpages don’t load. Windows itself is buggy.

 

[sketcheyes]

Update: Forgot to add this, not sure if this is part of the problem or not but my mouse is getting stuck now. I had this issue before but I guess it's come back now. For example, I will be playing an FPS and when I try to swipe right sometimes, it'll be as if my mouse is stuck and cannot be moved for a solid second or two. Happens on desktop too, almost as if it is locked in place for a second.

 

[ubuysa]

Does it BSOD in Safe Mode? I'd like you to try starting Windows in Safe Mode. In Safe Mode a stripped-down Windows system is loaded, with only critical services and drivers loaded. Typically no third-party drivers are loaded. This does mean that you won't be able to do any useful work in Safe Mode, or play games, and many of your devices may not work properly (or at all) because their drivers have not been loaded. Your display will be low resolution for example, because you'll be using only the Windows basic display driver.

The usefulness of Safe Mode is that because it's a stripped-down system consisting only of Microsoft services and drivers it's very stable, so if you get BSODs or crashes in Safe Mode you have a hardware problem. On the other hand, if it's stable in Safe Mode then your problem is with a third-party driver or service that wasn't loaded in Safe Mode. There is another technique we can use in that case to locate the problem service or driver.

 

[sketcheyes]

So I’ve been in safe mode for about 2 hours now, no BSODs yet. Switched back to a normal boot and blue screened right at the sign in page (which usually happens). Also able to run sfc /scannow in safe mode, not able to in a normal boot with turbo boost enabled

 

[ubuysa]

That does indicate that your key hardware is working OK, though it doesn't test turbo mode. I'd like you now to do a number of clean boots of Windows. First disable all startup apps and all third-party services, that should be stable since Safe Mode was. Note that a clean boot loads a complete Windows system with all Microsoft drivers so it is different from Safe Mode in that respect.

Then, leave all startup apps disabled and enable half of the third-party services, then reboot. If it BSODs the problem is in the half you enabled, if it doesn't BSOD it's in the half you didn't enable.

Working with the suspect half, enable (or disable) half of those and reboot. Working with half the services at a time you can soon zoom in on the problem one or two. This technique is called a binary search technique, it's very powerful and can be used in many scenarios.

If it doesn't BSOD with all third-party services enabled then do a binary search on the startup apps.

See whether that allows you to locate a faulty service or app. If not, don't worry, there is another technique we can use to try and discover what's causing the BSODs.

 

[sketcheyes]

That does indicate that your key hardware is working OK, though it doesn't test turbo mode. I'd like you now to do a number of clean boots of Windows. First disable all startup apps and all third-party services, that should be stable since Safe Mode was. Note that a clean boot loads a complete Windows system with all Microsoft drivers so it is different from Safe Mode in that respect.

Then, leave all startup apps disabled and enable half of the third-party services, then reboot. If it BSODs the problem is in the half you enabled, if it doesn't BSOD it's in the half you didn't enable.

Working with the suspect half, enable (or disable) half of those and reboot. Working with half the services at a time you can soon zoom in on the problem one or two. This technique is called a binary search technique, it's very powerful and can be used in many scenarios.

If it doesn't BSOD with all third-party services enabled then do a binary search on the startup apps.

See whether that allows you to locate a faulty service or app. If not, don't worry, there is another technique we can use to try and discover what's causing the BSODs.

BSOD during the clean boot.. Didn't even make it to the binary search part :/ .. I'm guessing it is a hardware problem then?

 

[ubuysa]

That might be due to a bad driver that's loaded at boot time. Try booting in Safe Mode and disable all third-party services and startup apps there. Then do a clean boot.

 

[sketcheyes]

That might be due to a bad driver that's loaded at boot time. Try booting in Safe Mode and disable all third-party services and startup apps there. Then do a clean boot.

Still BSODing. (Thank you for continuously helping by the way)

 

[ubuysa]

If it BSODs in Safe Mode then it's almost certainly a hardware problem, that's why we did this earlier.

Use another PC and download a Linux distro to a USB drive (Mint is pretty decent). They can run directly off the USB drive without needing installation. Run a Linux distro off the USB drive and see whether it's stable. Work the Linux distro hard, use all the devices you can to try and make it fail.

If it really is a hardware problem then the Linux system will crash too.

 

[sketcheyes]

If it BSODs in Safe Mode then it's almost certainly a hardware problem, that's why we did this earlier.

Use another PC and download a Linux distro to a USB drive (Mint is pretty decent). They can run directly off the USB drive without needing installation. Run a Linux distro off the USB drive and see whether it's stable. Work the Linux distro hard, use all the devices you can to try and make it fail.

If it really is a hardware problem then the Linux system will crash too.

Oh well it does not BSOD in safe mode. It did after I went into safe mode and disabled all third party apps and then did a clean boot after. (With turbo boost on)

 

[ubuysa]

Ah ok. Is it possible to boot into Safe Mode with turbo boost on? What happens if you can do that?

If it's not possible to boot into Safe Mode with turbo boost on then can you clean boot it, with all third-party services and startup apps disabled and with turbo boost OFF? If it's stable then try the same clean boot, with all third-party services and startup apps disabled and with turbo boost ON. If the first is stable and the second BSODs then I think you'll have pretty much confirmed it's the CPU.

 

[sketcheyes]

Ah ok. Is it possible to boot into Safe Mode with turbo boost on? What happens if you can do that?

If it's not possible to boot into Safe Mode with turbo boost on then can you clean boot it, with all third-party services and startup apps disabled and with turbo boost OFF? If it's stable then try the same clean boot, with all third-party services and startup apps disabled and with turbo boost ON. If the first is stable and the second BSODs then I think you'll have pretty much confirmed it's the CPU.

Ok so:

With turbo boost on it does not BSOD in safe mode. Does during clean boot.
With it off it never BSODs at all.

Also I managed to somehow managed to run a stress test with Turbo Boost on before my PC could crash and here is a screenshot of it failing at the "math" portion. Not sure what it means. Also CPU temp reached a max of 107 degrees Celsius during it.

https://prnt.sc/NOqkx45lw99Y

 

[ubuysa]

I think you definitely have a bad CPU. That may be down to the known issue with the i9-13900K (and i9-14900K) but 107°C is well over the 100°C max operating temp specified by Intel. I hope you have a decent cooler on there? If not get a decent cooler before you replace the CPU. I always over-spec my CPU cooler.

The problem with the known flaw is that once an affected CPU has been cooked it's dead and cannot be recovered. Intel are offering an RMA on affected CPUs. If this PC was bought from a shop or was custom built by a system integrator then you need to contact the PC supplier to get the CPU RMA'd. If this was a home build then you need to contact whoever you bought the CPU from. Failing all that you'll need to contact Intel directly (you'll need proof of purchase of the CPU).

Intel have released two microcode updates (x125 and x129) that will prevent affected CPUs from cooking themselves (but they won't fix an already cooked CPU). These are shipped as BIOS updates, you need to apply the lates BIOS updates for your motherboard (cotaining these microcode updates) to avoid cooking the replacement CPU.

 

[sketcheyes]

I think you definitely have a bad CPU. That may be down to the known issue with the i9-13900K (and i9-14900K) but 107°C is well over the 100°C max operating temp specified by Intel. I hope you have a decent cooler on there? If not get a decent cooler before you replace the CPU. I always over-spec my CPU cooler.

The problem with the known flaw is that once an affected CPU has been cooked it's dead and cannot be recovered. Intel are offering an RMA on affected CPUs. If this PC was bought from a shop or was custom built by a system integrator then you need to contact the PC supplier to get the CPU RMA'd. If this was a home build then you need to contact whoever you bought the CPU from. Failing all that you'll need to contact Intel directly (you'll need proof of purchase of the CPU).

Intel have released two microcode updates (x125 and x129) that will prevent affected CPUs from cooking themselves (but they won't fix an already cooked CPU). These are shipped as BIOS updates, you need to apply the lates BIOS updates for your motherboard (cotaining these microcode updates) to avoid cooking the replacement CPU.

After cleaning my fans and inside of the case, I now do not blue screen and am able to be on the computer however my Windows itself is very buggy. With turbo boost on, Sfc scans don’t work, chkdsk commands don’t work, downloads/installs don’t go through. When I tried to reset my pc, (left turbo boost on) my windows installation also failed. After turning it off, the installation went through. Also the Commands work with turbo boost off though. So I’m just wondering before I contact intel about my cpu, if this is still a 100% cpu problem? Thank you once again

 

[ubuysa]

I can't give you a guarantee, but since turbo boot causes all your issue then I'd say it's a fair bet that your CPU is bad.