Millions of physical CPU core issues following Intel Core I7 14700K undervolting
Millions of physical CPU core issues following Intel Core I7 14700K undervolting
For the non-kneecapped old microcode, verify the CPU voltage or adjust it (typically to 1.45–1.5V, though I'm not confident with Intel's latest chips). On a non-OC B660 board, it should likely be fine.
You're asking about the potential dangers of undervolting, comparing it to overclocking. It's a risky move that could cause issues if calculations go wrong. I'm worried about accidental file damage, like corrupting your Windows installation. Many people treat undervolting as a safe option because they're unfamiliar with the process, assuming it won't cause problems. They just lower the voltage and hope everything runs smoothly. This practice used to be more common, but now it's less understood. I think some online influencers contributed to the misconception. If you're unsure, it's best to revert your settings to the original voltage and ensure no data loss occurs.
From what you're observing in the screen, it seems core 8 is underperforming compared to others. This assessment is based on what I can see, though I'm not sure if other cores are also behaving normally. It's clear that certain conditions are needed for the system to operate at full capacity under heavy demand. Even though high temperatures are still within acceptable ranges, it suggests the chip is reaching its limits depending on voltage, cooling, and other factors. It appears you're likely operating near the chip's maximum capacity given these parameters. Comparing your results with others using different setups can help identify areas for improvement. Remember, OCCT will stress the machine fully in a real test, so unexpected errors are expected. You can review CPU voltage usage, speeds achieved, cooling methods, and use this data to refine your approach.
Yep, I understand that undervolting isn't that simple and stability issues on low voltages are inevitable, but I just found that weird that most of the people just get BSODs/WHEA Errors in OCCT, but I get this weird result with millions of errors. Probably as was said before by PDifolco, the processor is just on the verge of the complete failure because the supplied voltage is not enough (And now the affected core is just a random number generator), but it isn't quite there. Anyway, As I understand these exact CPU Core errors in that amount is a rare sight after undervolting, the consequences are mostly different and less scary) Thanks for your concern though, everything with the Windows 10 on which I have been testing the processor is fine, it is anyway a second operating system installed for testing purposes after I recently upgraded my PC
Errors initially show up only on this core at this stage, but increasingly more extreme undervolt settings also trigger issues. These are mainly weak cores; I can't even fine-tune their voltage individually. I'm considering older Microcode to see if it helps with Intel XTU, but it won't solve the problem. It seems these are efficiency cores designed for lower performance and multipliers. Regarding cooling, the tests suggest it works well despite some high-end coolers and expensive options. I'll verify the voltage levels from the tests as well. I assume the actual voltage might vary due to the motherboard, BIOS, BIOS settings, and the Microcode used. Therefore, stock voltages could differ, leading to varying results. The CPU performance is strong and above average, as confirmed by the Z-score, OCCT, and Cinebench scores.
It's great you have a test OS. Have you tried running without any undervolt? This chip comes from an era with hardware problems. I remember reading somewhere it was linked to the heatspreader bending, though others say it's not connected. You might want to experiment with a contact frame instead of the Intel clamp.
Everything is working smoothly without any issues. The system feels solid and everything looks good. The case is brand new, just picked up at the hardware store. I plan to add a contact frame later once I upgrade my case, power supply, and cooler—maybe it will help with heat management.
I observed that you considered the minor variations between different boards, BIOS versions, and similar factors, indicating a solid understanding of the system's behavior. Default configurations remain consistent across models according to the specifications, though individual boards might vary slightly. Instability may also stem from other elements such as the board itself, RAM, or external influences impacting outcomes. Once you implement a more effective cooling setup, you can attempt the test again with your existing settings to determine if improvements are possible. Keep in mind the device's inherent limits—if it's already reaching its maximum capacity, that’s the point. There are potential workarounds in certain situations, mainly involving adjustments, so continue experimenting and you might resolve the challenges.