The story of 13th and 14th generation processors and my remaining questions

[sMyAsiN]

Hello to all Intel fans and enthusiasts ,

I have some questions about the 13th and 14th generation processors (600 and 700 series) regarding the problems and horror stories of the last few years or months and I will be happy to answer my questions and discuss the old hot discussions once again! ; (If I was not good at explaining, I apologize to you).

The story started when Intel introduced its new architecture, its 12th generation 10nm, to the market! Very different architecture and hybrid cores and DDR5 RAM! Which provided extraordinary performance for users; At first, it suffered from compatibility problems and inconsistencies between the processor and the operating system and other components, and over time, it improved greatly with frequent updates to the operating system and new BIOSes; As the process of improvements and improvements continues, we reach the 13th and 14th generations, where the main problems of the 13th and 14th generations were and are extremely high power and extremely wild heat. In the early production of the 13th generation, there was an oxidation problem, which Intel said would be fixed in 2023! A physical problem that could not be solved with a firmware update and was in the manufacturing process and perhaps casting (maybe something like that!) that caused the transistor insulation to oxidize and eventually, over time, caused the processor to fail or completely fail; fortunately, this issue did not exist in the 14th generation processors!

Finally, the big problem that later became very surprising, annoying, and very complicated was the problem of the failure of the 13th and 14th generation processors, which after a long time or so, due to faulty microcode and incorrect motherboard BIOS settings, led to the destruction of the Ring Bus of the cores and ultimately to the crashing or permanent failure of the processor over time. Now it was not clear exactly where the problem came from! Was the quality of the silicon in the processors not high enough? Or had the motherboards gone too far and not followed the correct instructions? Or maybe Intel really or "intentionally" wanted it this way! (After hiding the oxidation problems and confirming it with a long delay, I feel that Intel is intentionally the main culprit (in recent generations of motherboards, the PL1 and PL2 limiters were not very important and were mostly on automatic or Unlimit mode)); During this vulnerability period, issues were expressed to compensate for or prevent this risk until Intel solved the problem; I had heard that most of this problem was affected by SKU K/KF/KS and the cause was the request and increase of abnormal and excessive voltages during high pressure and temperature, which caused the destruction of the processors, or the lack of optimization or compatibility of DDR5 RAM with very high frequencies caused this problem, or the excessive frequency of the processors caused a severe reduction in their life, and some pointed to the low-power E cores, which caused premature corrosion and destruction of the processors. . !; However, most of the problems were affected by high frequencies and high voltage requests, but I saw and heard that someone in the processor The 14900k was working perfectly fine, away from high pressure, high temperature, high voltage and high frequency, in xmp mode off and keeping its BIOS updated, but eventually after months its processor lost its quality over time and over time its processor had gone through the process of destruction!, and I guess that person was unlucky and had a poor quality processor! No one knows this as much as Intel! - I have not heard of the K/KF series i7 and i5 processors, at least I have not heard of the processor failing permanently and I have only heard of the problem of them crashing, which was resolved by updating the BIOS! By the way, why were the non-K 13th and 14th generation Intel processors not affected? It is true that they have low consumption, but in this case I must say that as many have pointed out and we have seen, the problem of processor consumption was not even the one who severely limited his 13th and 14th generation K/kf processors and corrected them by DownClock or Undervolt and Or weakened its power, over time it would have led to destruction! Can't K-series processors withstand high pressure and temperature? So how are non-K processors not affected? They could consume at least 150 watts! And compared to the weakened or undervolted and downclocked K-series processor that consumed the same as non-K, they failed? I guess they never reached or could not reach dangerous voltages.
A few other things have been interesting to me; I have heard that the quality of processors can be guessed from its batch number (FPO)! Is such a thing possible? I had heard that processors whose FPO starts with L means that they were packaged or manufactured in Malaysia and a processor that starts with a batch code X is often from the US and is of higher quality! And the higher the next digits, the newer and better quality the processor is! Is the speculation correct? If so, it is really funny! Why should there be a difference in quality between X and L and others Fabs? Don't they all follow the same guidelines? It might make sense that newer processors would have the same and better quality, but naturally, processors may not be of the same quality when packaged, which means they should be retired! How is it that the processor is identified by its batch number (FPO) and there may be a difference in its quality? Is it possible to find out whether at least our 13th and 14th generation processors are affected by the degradation or not? Is there any clue to how vulnerable the processor is or what quality of silicon it is made of?
It is interesting that according to the channel YouTube "Level 1 Techs" problems with i9 series processors used in servers after they failed and were replaced; He also said that the processors were away from a harsh environment and with the necessary precautions against high voltage, and temperature, and even using processors below the frequency of 4800 DDR5 RAM caused the destruction of the processors over time! He said that the motherboard used was the W680, which did not have the aggressive behavior of the Z series motherboards, and that the W680 motherboard was simply an optimized and more conservative motherboard designed specifically for a specific range of Intel's expectations, which are separate from any excessive pressure and additional power to the processor! However, he reported that the processors were affected by the destruction!! My main and key question is whether the B600 and B700 and H600 and H700 series motherboards also affected the processors or not? This is the most important question I want to know the answer to because we know that Z motherboards are often more advanced and more rugged, they have very powerful features They have a lot of complexity in their right that the B and H series motherboards don't have and they are much more limited, so their risk can be less right?! When Intel recommended to keep eTVB on, I was looking for it in the BIOS of my B760F motherboard and I never found it! And in fact it didn't exist or maybe it was hidden without people having access to it! And I don't know how I could have protected myself from the vulnerability of processors when the microcode was not yet released! . I have not received any reports that using processors on B and H chipset motherboards caused them to crash and I have not seen anyone even crash and witness their processor being destroyed; is this really the case? Are most of the crashes on Z chipsets? Or is it not relevant and it is possible that the B and H series chipsets can damage the processor! What is the reason and reality of the 13th and 14th generations? Why and how does this happen? Who is right and is it still Is there anything we're missing?

Thank you for reading this very long article! I really appreciate your responses and comments, friends .

[PC1997]

Hello! As an enthusiast, I’d be happy to answer some of your questions and offer clarification.

First, some basics: voltage (V) is often referred to as 'electrical pressure' (as you elegantly put it) that pushes charges through a circuit. Higher voltage means more push. However, a more accurate definition is that voltage is a potential difference that creates an electric field, which then exerts a force on electrons.

Amps, short for Amperes, measure the electric current—the rate at which electric charge flows through a circuit. 1 Ampere equals 1 Coulomb of charge passing a point per second. Since 1 electron carries approximately 1.602 x 10⁻¹⁹ Coulombs, 1 Amp is approximately 6.242 x 10¹⁸ electrons passing a point per second.


Ohm’s Law is a fundamental principle in electrical engineering and physics that explains the relationship between voltage (V), current (I), and resistance (R) in an electrical circuit.

The law states: V = I x R

Where:

•  V is the voltage across the component (in volts, V)
•  I is the current flowing through the component (in amperes, A)
•  R is the resistance of the component (in ohms, Ω)

Raptor Lake CPUs are prone to degradation due to vulnerabilities in the clock tree circuit of the IA core under elevated voltage and temperature. Microcode updates are intended to compensate for the physical design limitations of the chip and help mitigate degradation effects that can’t be physically altered.

Furthermore, motherboard vendors were releasing overly aggressive default BIOS power delivery profiles. To summarize what you already discussed, too much voltage and excessive current were being delivered to the CPUs causing them to degrade more quickly.

Also, once the CPU begins to degrade, it becomes increasingly sensitive to temperature—even within its thermal design limits—leading to instability. This often necessitates reducing clock speeds to maintain thermal and operational stability.  Eventually, if the degradation is allowed to progress long enough, the CPU will no longer function correctly at any reasonable temperature or stop functioning entirely.

Regarding temperature, it’s important to state the obvious: if everything else is equal, a CPU that runs at 50°C will last longer than one that runs at 100°C. That said, the primary cause of degradation is elevated voltage and current, rather than high temperature alone.

Ironically, better cooling can allow higher power draw, potentially accelerating degradation due to increased voltage and current. With weaker cooling, degradation takes longer since you can't push as much power without constantly hitting the TJ (Thermal Junction) limit.

And while motherboard vendors have gotten way better at their default power delivery profiles, there is still no agreed upon standard and most of them still do whatever they think will sell their motherboards over the competition. The real issue isn’t that vendors create these profiles—it's that they’re enabled by default.

As far as the B series motherboards (I'm mostly familiar with the Z series) but my guess would be simply the aforementioned power delivery settings discussed is the only reason you don't see more rapid issues, but with older microcodes the CPU would still degrade, even on those boards.

On the subject of batch numbers... batch numbers mainly matter for overclocking potential. While some batches may perform better, stock performance varies less and comes down mostly to luck. There is variation even among the KS SKU. Between the two 13900KS and 14900KS CPUs I own, the former undervolts more efficiently, while the latter overclocks higher but requires more voltage.

On the subject of DDR5 memory that you mentioned: each generation CPU has limits to what speed and capacity they can handle while maintaining stability. So the more RAM capacity and speed you have, the more demands will be put on the memory controller on board the CPU.

In the end, the engineering teams that made the most mistakes may be best suited to lead—because they’ve learned the most.

[sMyAsiN]

@PC1997  oh sorry , I forgot to click Reply .

[sMyAsiN]

Hi @PC1997,

First of all, I sincerely thank you for your honest answers to my questions.
After reading your replies, a few more questions came to my mind:

You mostly attributed CPU degradation to voltage. I’ve heard that for CPUs like the 13900K and 14900K, to stay safe, the voltage should not exceed 1.5 or 1.55 volts. Is that true?
Does this refer to sustained voltage over time, or does it include peak (momentary) voltage as well?

As for i7 K and i5 K series CPUs, I haven’t heard anything specific about safe voltage limits.
Can we assume that 1.5 to 1.55V is also the maximum safe range for those CPUs?

I initially tried to increase current instead of voltage.
However, I found that when reducing voltage, the frequency also dropped significantly.
At first, I undervolted and underclocked my CPU heavily, thinking this would help.
But the CPU also consumed less current in direct proportion to the voltage drop! ("offset /adaptive - Mode" ~ "-0.055" to "-0.075")

And that wasn’t the only issue. With low voltage, the CPU became unstable,
and sometimes, before instability occurred, the motherboard would compensate by boosting voltage,
causing idle or low-load voltages to rise even higher than before undervolting.
(The CPU I’m using is a 14700KF.)

Instead of changing CPU Core Voltage, Global Core Voltage SVID, or Actual VRM Core Voltage
(in motherboards like B760-F and Z790 MASTER — I'm more familiar with the STRIX BIOS),
I used the IA VR Voltage Limit option to ensure the voltage would not exceed 1.5V,
since I don’t really trust Intel’s safeguards.

(My 14700KF, in stock settings, didn’t go above 1.48V anyway — at least according to software readings,
since I don’t have an oscilloscope!.)

Still, I used options like IVR Transmitter VDDQ Voltage,
IA AC Load Line & IA DC Load Line (set around 1.5 mOhms, above the default 1.1),
and turned IA CEP = ON.

I found these options to be more reliable,
especially because I wasn’t sure about the exact PL1 and PL2 values at first.
So I experimented using all of these settings.

Despite calculating based on the power formula (W = V × I) , resistance, and voltage,
I couldn’t achieve my desired parameters and ended up constantly dealing with reduced CPU performance.
All I could do was wait and hope that Intel would eventually solve these issues.

Another question that’s bothering me is this:
How come some CPUs((only 14th Gen) like ,14900k) , still degraded or got damaged,
even though they were used in safe and well-maintained conditions,
with reduced performance, low power usage, and carefully limited voltages/currents?

Especially those CPUs that weren’t even part of the 13th generation?

Now, I have a few short questions that I’d appreciate quick answers to:

Isn’t it extremely dangerous when CPU temperatures exceed 100°C, like up to 115°C?
Intel's specs for previous-gen Core Ultra desktop CPUs (Arrow Lake) say the maximum temperature limit is 100°C.
Even 100°C already seems high. So why risk going beyond that?

What’s the actual thermal threshold at which Intel CPUs burn out?

Early on, because of Intel’s instability issues, the risk of CPU degradation,
and very high temperatures reaching 100°C,
I used the Asus Multicore Enhancement Limit at 90°C option
to prevent my CPU from reaching dangerous temps.

Does this setting only filter/limit the temperature,
or does it also affect CPU power consumption?

Is this option effective on motherboards that lack more advanced temperature limiting controls,
and does it help prevent long-term CPU damage?

Finally ,
In Asus BIOS, what does the option Synch ACDC Loadline with VRM Calibration actually do?
Does it synchronize IA AC Load Line and IA DC Load Line?

Where is this useful?
(I’ve rarely seen anyone talk about it!)

Also, does adjusting LLT (from levels 1 to 7 depending on the motherboard),
or in Gigabyte boards (from Normal to Ultra Extreme),
cause the voltage to exceed the VID limits of the CPU?

Thanks again — like always,
Thanks Steve .

[PC1997]

You mostly attributed CPU degradation to voltage. I’ve heard that for CPUs like the 13900K and 14900K, to stay safe, the voltage should not exceed 1.5 or 1.55 volts. Is that true?

Yes, you definitely do not want to sustain or push those voltages for extended periods. Short bursts, such as transient loads, are generally fine. (Note that you won’t be able to observe transient loads without an oscilloscope.)
Software monitoring tools like HWiNFO—an excellent application—can provide voltage readings, but, like all software, they aren't always perfectly accurate. Their accuracy depends on the sensors providing the data.
For the most accurate voltage readings, it's best to measure directly on the motherboard using a high-precision voltmeter.
The latest microcode does a great job of capping voltages at 1.55V, but this only applies when using Intel’s default settings. If you enable overclocking profiles, the motherboard can still override these limits.

Does this refer to sustained voltage over time, or does it include peak (momentary) voltage as well?

The lower the sustained voltage—while still maintaining system stability—the better. In fact, the first CPU cores to degrade are almost always the two best-performing P-Cores, which are used by Thermal Velocity Boost to reach the processor's maximum advertised frequency. Under light loads, if the die temperature is low enough, these cores will boost to maximum frequency for short periods. Doing so requires higher voltages to maintain stability, which in turn causes these cores to degrade faster.

This issue should already be addressed with the latest microcode update, but it may still pose a long-term concern. Syncing all your CPU cores can help avoid this problem, as it prevents uneven stress on individual cores. However, this won't stop the CPU from degrading if the voltage and current (amperage) remain too high for extended periods.

Regardless of configuration, if voltage and current levels are excessive over time, CPU degradation is inevitable. The key question, then, becomes: at what frequency—and therefore, at what required voltage to maintain stability under varying load conditions—will the current flowing through the CPU contribute to its degradation.

As for i7 K and i5 K series CPUs, I haven’t heard anything specific about safe voltage limits.
Can we assume that 1.5 to 1.55V is also the maximum safe range for those CPUs?

Intel specified a maximum safe voltage of 1.72V! Of course, I assume they mean peak, short bursts—but I’ll leave that to Intel to clarify. In my opinion, 1.5 to 1.55V is more realistic as the maximum, though even that shouldn't be considered safe for sustained use.

I initially tried to increase current instead of voltage.
However, I found that when reducing voltage, the frequency also dropped significantly.
At first, I undervolted and underclocked my CPU heavily, thinking this would help.
But the CPU also consumed less current in direct proportion to the voltage drop! ("offset /adaptive - Mode" ~ "-0.055" to "-0.075")
And that wasn’t the only issue. With low voltage, the CPU became unstable,
and sometimes, before instability occurred, the motherboard would compensate by boosting voltage,
causing idle or low-load voltages to rise even higher than before undervolting.
(The CPU I’m using is a 14700KF.)

Thanks for the detailed explanation—what you're experiencing is a common but nuanced set of behaviors in modern CPUs, especially with Intel’s recent chips like the i7-14700KF. Let’s unpack what's going on and clarify why your initial undervolting approach produced those side effects.

Key Concepts in Undervolting and Adaptive Behavior

1. Undervolting via Offset/Adaptive Mode

You're using offset/adaptive undervolting (e.g., -0.055V to -0.075V). This tells the CPU to reduce voltage relative to its VID (voltage ID), which is dynamically adjusted by the motherboard/BIOS based on load, temperature, power limits, and other telemetry.

• Offset mode just blindly subtracts from the VID.

• Adaptive mode allows VID to scale based on workload, but your offset still applies.

At idle or light load, your undervolt might work fine. Under load, the VID might spike or the chip might demand more power than you expected—leading to instability.


Why Frequency Drops with Voltage
Modern CPUs use DVFS (Dynamic Voltage and Frequency Scaling). If voltage drops below what's needed to maintain a target frequency:

• Either the system crashes,

• Or the motherboard/CPU firmware throttles the frequency to stay stable.

This explains your observation of lower frequency under heavy undervolt.

Current Decreases with Voltage

You noticed this:

"CPU also consumed less current in direct proportion to the voltage drop"

That's expected from Ohm’s Law and power equations:

• Power = Voltage × Current

• Lower voltage → lower power (if frequency also drops) → lower current drawn

But that’s not always desirable! If your goal is efficiency without major performance loss, this becomes a delicate balancing act.


Motherboard Voltage Compensation

You saw this behavior:

"Motherboard would compensate by boosting voltage, causing idle or low-load voltages to rise even higher"

That’s typical of modern motherboards using Load-Line Calibration (LLC) and voltage guard bands. They detect instability or overly aggressive undervolt and then:

• Increase VID temporarily to stabilize

• Even overcompensate at idle, causing higher idle voltages

Why You Can't Just Drop Voltage Aggressively

With a CPU like the 14700KF, Intel's boost algorithms (Turbo Boost 2.0/3.0, TVB, etc.) are complex and tightly linked to:

• Temperature

• Power consumption (PL1/PL2/Tau)

• Current limits (ICCmax)

Undervolting too aggressively interferes with these controls. The CPU or motherboard firmware reacts by trying to “correct” what it sees as a potentially unstable or inefficient setup.

Recommendations

• Use Intel XTU or BIOS Adaptive Voltage Carefully

• Stick to modest offsets like -0.030V to -0.050V at most.

• Combine with manual tuning of LLC to avoid overshoot at idle.

• Monitor Frequency Scaling Behavior

• Use HWInfo or Intel XTU to track effective clock, VID, and power draw.

• Ensure frequency isn’t dropping excessively under load.

• Avoid Heavy Undervolts Without Testing Stability

• Use tools like OCCT, Prime95 (small FFTs), or Cinebench R23 loop for stability testing.

• Consider Setting Power Limits Instead

• Instead of undervolting, set PL1/PL2/Tau to more efficient levels (e.g., 150W instead of 253W).

• This forces the CPU to limit power without relying purely on undervolting.

• Tune LLC (Load-Line Calibration)

• Try LLC level 4 or 5 (varies by board) for better control of Vdroop under load.


Part 1: BIOS Settings (Z790 Strix)

These give you the most control and stability.

Goal:

• Safe undervolt with minimal performance loss

• Avoid high idle voltages and instability

• Moderate temperatures and power draw

Recommended BIOS Settings:

1. AI Tweaker Tab:

AI Overclock Tuner: XMP I or Manual

CPU Core Ratio: Auto or Sync All Cores (optional: per P/E core tuning)

Performance Core Ratio: Auto (or fixed if overclocking)

BCLK Frequency: 100.00

ASUS MultiCore Enhancement: Disabled - Enforce All Limits


SVID Behavior: Typical or Best-Case Scenario

CPU Core/Cache Voltage: Adaptive Mode

Additional Turbo Mode CPU Core Voltage: 1.25V (baseline for stability testing)

Offset Mode Sign: -negative

CPU Core Voltage Offset: Start with -0.040V

IA AC/DC Loadline: 1 (optional, for more aggressive undervolting stability)

2. DIGI+ VRM:

Load-Line Calibration: Level 4 or 5 (helps avoid overshoot at idle and droop at load)

Current Capability: 140–170% (allows VRM headroom, optional)

3. Advanced > CPU Configuration:

Intel SpeedStep / Speed Shift: Enabled

Turbo Mode: Enabled

Thermal Velocity Boost: Enabled

TVB Voltage Optimizations: Enabled (optional)

4. Advanced > Platform Misc Configuration:

C-States: Enabled

Package C-State Limit: C7 (for better idle power saving)

Save & Test Stability using OCCT or Cinebench R23 for at least 10–15 min loops at full load.


Part 2: Intel XTU Settings

Intel Extreme Tuning Utility (XTU) is a safer, Windows-based way to tune voltage and clocks—good for experimenting before hard-coding into BIOS.

Steps in Intel XTU:

1. Download & Install

• Download from Intel’s official site

• Launch as Admin

2. Basic XTU Undervolt Settings:

Core Voltage Offset: -0.040V to start (go in small steps: -0.010V at a time)

Turbo Boost Short/Long Power Max: Optional: reduce to ~180W for better thermals

Turbo Boost Power Time Window: ~56s or less

Turbo Boost Ratio Limits: Leave default unless overclocking

3. Advanced Tuning Tab (optional):

• Adjust E-Core and P-Core offsets separately if needed

• Monitor effective frequency, thermal throttling, and current limit throttling

Stability Testing (XTU or 3rd Party):

• Start with Cinebench R23 (multi-core loop, 10–30 min)

• Then use OCCT or Prime95 (small FFTs) for ~30 min

• Watch for:

• Crashes

• WHEA errors (check Event Viewer)

• Clock throttling (monitor in HWInfo)

Your Targets

Idle Voltage: 0.65V – 1.05V

Load Voltage (all-core): 1.15V – 1.25V

All-core frequency: ~5.3 GHz (P-Cores), ~4.2 GHz (E-Cores)

Package Power: 125–253W (depending on limit set)

Temps under load: < 90°C ideal (spikes OK)

Final Advice

• Don’t chase the lowest voltage—chase stable efficiency.

• Once you're happy in XTU, replicate the settings in BIOS for permanent application.


Isn’t it extremely dangerous when CPU temperatures exceed 100°C, like up to 115°C?

Yes! Running a Raptor Lake—or any modern CPU—at 115°C is a guaranteed way to accelerate degradation. 115°C is basically Chernobyl for your processor.

What’s the actual thermal threshold at which Intel CPUs burn out?

I'm not sure, but check out this excellent YouTube video by Asianometry:

"Why the Chips Get Hot" – https://youtu.be/US6YO-IK64w?feature=shared

Early on, because of Intel’s instability issues, the risk of CPU degradation,
and very high temperatures reaching 100°C,
I used the Asus Multicore Enhancement Limit at 90°C option
to prevent my CPU from reaching dangerous temps.
Does this setting only filter/limit the temperature,
or does it also affect CPU power consumption?

Reducing the maximum operating temperature also reduces power consumption—unless your cooling solution is efficient enough to prevent the system from reaching those high temperatures in the first place. If the system is thermal throttling, then yes, lowering the temperature will reduce power consumption.

In Asus BIOS, what does the option Synch ACDC Loadline with VRM Calibration actually do?
Does it synchronize IA AC Load Line and IA DC Load Line?

Yes — it does synchronize IA AC Load Line and IA DC Load Line with the behavior of VRM Loadline Calibration, aiming to ensure better accuracy in power delivery and monitoring.

Where is this useful?
(I’ve rarely seen anyone talk about it!)

When to Enable:

Useful if you're using manual LLC settings and want power reporting and thermal throttling to align better with real behavior.

Recommended when overclocking, as it ensures consistent performance and avoids over- or under-reporting CPU power draw.

When to Disable:

If you're manually tuning IA AC/DC Load Lines separately for fine-grained power control.

In certain scenarios where you want the CPU to believe it’s drawing more/less power than it is (e.g., certain undervolting strategies).

Also, does adjusting LLT (from levels 1 to 7 depending on the motherboard),
or in Gigabyte boards (from Normal to Ultra Extreme),
cause the voltage to exceed the VID limits of the CPU?

Yes, adjusting LLC (Load-Line Calibration, often miswritten as LLT) on motherboards — including Gigabyte's modes like Normal, High, Turbo, Extreme, Ultra Extreme — can cause the voltage to exceed the CPU's VID (Voltage Identification Definition) limits, especially under load or idle conditions, depending on how aggressively it is set.

[sMyAsiN]

Thank you very much for answering my questions once again.

I’ve learned a lot of useful information from you, and I truly appreciate it.

• I often watch videos from the Actually Hardcore Overclocking YouTube channel,
  but sometimes I get confused.
  
  Apologies for any spelling or phrasing mistakes.
  
  

Recently, I updated to microcode 0x12F (though I’m not sure whether this microcode has any issues or bugs).
I followed your advice and managed to partially reach my goals.
I hope this new microcode has a positive impact and resolves most — if not all — of the current problems,
though future updates may still be necessary.

With the current settings, I’ve managed to achieve the following on my 14700KF processor:

• PL2 power limit: 188W
• ICCMAX: 249A
• Under heavy stress testing or rendering,
• Voltage under full load: between 0.995V and 1.050V
• LLC level: 3
• IA AC/DC Load Line: set to 1.1 mOhms (Auto Mode)
• All cores fully loaded, power consumption maxed
• Finally, latest CPUID version score: I got ~13,000 Multicore
  (That’s actually slightly better than the maxed-out default settings for a 12900K!)
• Most importantly, the maximum temperature remained stable at just 64°C !.

This performance is clearly lower than Intel’s stock configuration,
but what matters to me is protecting my CPU from degradation or long-term damage
— at least until all major issues are fully resolved ,

Because RMA is difficult for me,
and since I’m using a tray CPU,
any physical failure would be very hard to replace..