I9--14900K CRC-Error while unpacking ZIP Files and Bluescreens - Hot CPU

[Gradye]

My System:

Gigabyte Z790 Aorus Elite AX ICE (BIOS F3)

Memory 2 x 24 GB GSkill F5-7600 Ram (DDR5-7600 / PC5-60800 DDR5 SDRAM UDIMM)

CPU Intel Core i9-14900K

GIGABYTE GeForce RTX 2080 GAMING OC 8G (GV-N2080GAMING OC-8GC)

Watercooling Sharkoon S90 RGB AIO 360 mm

Behavior:

At the beginning I still have a benchmark score with the Intel Extreme Tuning Tool in the standard settings of a little more than 8000 points, after I noticed the problems, I made the attempts to stabilize the system and get at least approximately suitable performance. Now I reach a maximum of 5000 points in the benchmark and notice that the system is unstable and even slower. I´m not only get CRC errors when unpacking ZIP files, but also regular bluescreens. Apps are closing without a message and the temperature is above the Limit, when idle, so that the power will reduce without a reason.

Tested:

I change the Thermal paste (Aairhut GX-14 Thermal Paste, 15.7W/m.k)

I have already tried several settings and adjusted the power limits. In the meantime I have switched off the INTEL Turbo function and still have an unstable system with not even close to the performance I paid for. The settings of ICCMAX, CEP, TVB from a Intel Chart. reduce even more performance, so that I am more comfortable on the road than I can say that I have a power CPU. Does anyone have any idea what I can do to get into the range I paid for?

[zzetta]

Hi. Try installing F1 BIOS, load defaults, then apply xmp and try again.

[Gradye]

Hello zzetta,

Thanks for the quick reply. I checked the Gigabyte website and the F1 BIOS is no longer available for download, but I did see that F4 is available for download. I will try this once and then test the system only with the power limit set, which should then exploit the potential up to the power limits without BSD, CRC error, closing apps or with temperatures that are not justifiable. That would at least be close to the performance I paid for. Any further intervention in the setting beyond automatic seems a bit excessive to me, especially for people who don't have that much background knowledge about computers, in order to fix an Intel problem. Even flashing the BIOS goes far beyond the knowledge of most customer.

[zzetta]

I have checked the motherboard page and there is an F1 Bios. Make sure you search for it carefully. The issue is not with Intel processors, but motherboard vendors changing AC loadlines without mentioning that in the patch notes of newer BIOS versions. I had a similar issue with my Gigabyte and rolling back to the bios from late august fixed everything. Try to give me the link for your motherboard. ALso make sure you select the correct Revision.

[Gradye]

Yes, I saw that too. I quickly checked and upon closer inspection realized that it wasn't the right board. The BIOS has been flashed. I only adjusted the power limits to 121/253. Everything else remained the same. After restarting, encountered several Blue Screens, including:

• Critical Process Died
• Attempt write to read-only memory
• wimfsf.sys

Only after disabling INTEL Boost was the system able to start. Attached are the BIOS, HWinfo, Intel Extreme Tuning Tool and other settings I found on the internet.

[zzetta]

I'm glad that you could flash to F1. Please reset the bios to defaults and don't touch anything else. Try again to see if it crashes with everything on default.

[Gradye]

Unfortunately, no success. I loaded the defaults and restarted the PC. Encountered several Blue Screens:

• 2x KMODE Exception not handled
• System Service exception

Then Windows started, but after a short time (10 sec) another Blue Screen:

• Unexpected Kernel Mode Trap

Upon the second Windows start, another Blue Screen (4 sec):

• Multiple IRP Complete requests

Restart (Blue Screen):

• System Service exception

Restart (Blue Screen):

• Clock watchdog timeout

Disabled Intel Turbo Mode and the system started. I restarted HWInfo and the Intel Tool.

[zzetta]

Unfortunate. Please reseat your ram, try with one stick at a time. You might have faulty ram. Unplug your PC and check your cables from psu to be inserted correctly. After you did that reset the bios.

[Gradye]

The temperatures constantly trigger core throttling. I tested the Intel Tool once with and once without power limits. I can rule out the memory issue since it has been tested in another system where the memory runs stably.

Sometimes you get, with fewer GHz than the maximum settings, a better performance in benchmarks because core throttling isn't triggered. I've had better experiences with this system when I set the P-cores to a maximum of 5 GHz and the E-cores to 4 GHz, but that shouldn't be the case considering the price of the CPU.

Now the Powerlimits are set to 121/253 with the result attached to this post.

[zzetta]

I was actually going to ask you about those high temps. Your cooler might not be working, or it is not mounted properly. Check the cables from pump, but you should first pull it out, reapply the paste and check if you did not forget to remove the plastic film from the pump block.

Also, regarding the ram, it's better to check if it's on the qvl of the motherboard.

[Gradye]

That was also my first thought. I deliberately used better thermal paste, as mentioned above. The pump is working; that has also been tested. I've tried everything possible. Since I work in IT, I initially tried to solve it myself, and only after realizing that I'm not alone in experiencing issues with Intel's high-end CPUs did I decide to describe my problem here. My main concern isn't necessarily finding an immediate solution; I can work with manual settings. But imagine the customers who don't have the competence to implement the suggestions here in the forum, let alone know what a BIOS is.

[zzetta]

You have abnormal cpu temps. I'm pretty sure your cooler is not mounted properly. Check for the things I suggested.

Yes, the cpu instability is a real issue, but first try to fix the cooling problem. You shouldn't have to limit your cpu to such drastic values. I have a 14700kf working with a Noctua u12a, a very small cooler and I'm able to continously hold 300w of power with no issues.

[Gradye]

I specifically purchased this cooler and tested it with the same results.

DeepCool AK400 Digital CPU Cooler

[zzetta]

Well, in this case, it might be possible that your cpu is physically busted. Contact Intel, or the local seller to give you a replacement.

But tbh, these weird temps tell me your cooling mechanism might be broken, or wrongly installed. I'm not trying to be disrespectful. Just my 2 cents.

[Gradye]

As I mentioned, that was also my initial thought. I've applied new thermal paste multiple times, even bought better paste afterwards. The idea that the water cooler isn't seated properly every time I try seems unlikely, as I always get similar results. I even tried tightening it by hand to prevent the CPU from bending and experimented with increasing the pressure by tightening it further, but it didn't really bring about any improvement, and I tried this with both fan versions. 

I guess I'll have to wait until Intel or Gigabyte come up with a solution that doesn't require BIOS intervention. I'm curious to see how they plan to roll it out, especially for those affected who don't have an IT background. It's unlikely they'll be easily persuaded to install a BIOS update or an Intel fix.

[zzetta]

Your temperature issue is definitely not caused by the bios or the settings. It's either a physical defect in the chip or the cooler itself.

[Gradye]

Those are the current temperature and GHz readings while idle with a YouTube video running.

[zzetta]

It's quite high given that you reduced the power limits and the frequency of the chip. Definitely not a normal behavior.

[Gradye]

in my opinion, especially after it was constantly at 100°C for the past two hours while we were testing what could help, it is normal. As I mentioned: 2 fans (AIO / Air). It's not that Intel CPUs can't handle high temperatures, but rather that they reach these temperatures very quickly without any apparent reason when Turbo Mode is activated, and then the CPU gets throttled. My screenshot shows that 5 GHz are not a problem for the P-cores without the temperature rising significantly. Now I'm getting better values than with the default settings from the BIOS, and the only things set are the power limits and Turbo Boost disabled.
The screenshot shows that all CPU cores are running at the same speed and are not being regulated, as far less power would be necessary for idle.

[zzetta]

Show me a picture of hwinfo when you are testing the cpu. Show me the cpu package power and the voltage. It's on the enhanced menu of your cpu in hwinfo.

From what I can see you have abnormally high temps with such low vids

[Gradye]

I had to switch HWInfo to English first. Are you referring to these values in HWInfo?