Question Overclocking i5 2500k beyond 4.5GHZ with a Hyper 212 EVO

Jul 31, 2020
2
0
10
0
I have been clocking my 2500k on P8Z68V at around 4.5GHZ at 1.3500V with a Hyper 212 EVO. Below are the snaps of the tweaks in the BIOS when at 4.5GHZ.



It idles at around 41-43C at 1.6GHZ however, when I tested P95 Tests at 4.5GHZ for 6 Hours without any issues. I got temps of about 75-77 C Avg during these tests.

When I was stress testing the CPU the voltages spiked to 1.360V from the base of 1.336 as well where as I have set the Voltage to 1.3500V. And I find the Temps or Voltage shows different by different applications like GPUZ and HW Monitor etc Temps Sensor or the Voltage Sensor.
My Question : Why is this happening even when I set the voltage to 1.3500V? If this based on the Speedstep Option that I have enabled in the Bios which one is correct? Volts of 1.368V, Are they in safe Range and can be used for daily usage?



But when I try to Increase the Multipler from 45 to 46 or 47 keeping at the same voltage i am getting BSOD with P95 tests.
My Question : So, Should I try increasing the voltage to 1.375 and then try increasing the multiplier to 47 or 48? Will these voltages damage the CPU or motherboard? Should I not try doing that? Are these voltages in safe Range for daily usage?

Your feedback will be highly appreciated and will be very helpful.
TIA:)
 

madmatt30

Titan
Ambassador
The voltage difference is down to llc overcompenstaing.

Honestly wouldn't try pushing further voltage or clock wise.

Thats a good overclock, youve likely hit the wall where temps and voltage increase aren't worth the minimal extra performance you'll gain.
 

Lutfij

Titan
Moderator
Those are some high voltages for a i5-2500k, I'd advise sticking to a 1.25~1.3v limit to retain the processor since people on OCN have seen their platforms conk out with higher voltages, although they did get some impressive clocks like you.

It's the very reason that different app's report different numbers that leads us to suggest cross referencing the numbers with BIOS and HWMonitor. I use HWMonitor since it's pretty much close to what BIOS is set to. You could also use Intel's XTU to manipulate the voltages.

Oh, right, latest BIOS versions also play a part in how the platform behaves with respect to overclocks and stability.
 
Jul 31, 2020
2
0
10
0
Thanks for the Reply @madmatt30

The voltage difference is down to llc overcompenstaing.
So, should I try lowering the Load Line Calibration?

Thats a good overclock, youve likely hit the wall where temps and voltage increase aren't worth the minimal extra performance you'll gain.
Yes, but these temps are Synthetic Burn Tests and will reach these temps with Any Video Encoding or Rendering or even the Highest CPU Intensive Games.

I did some video encoding before and after the overclock I gained a 3 Min time difference(almost) for 1.8G MP4. The Application was even using single core only. So, I think this is quite a difference in real world test.

When the CPU without any overclock works in TurboBoost Mode I found frequencies reaching 1.2900/1.3000 so thought increasing to 1.350 and test the different multiplier settings. Now its stabling at 1.35V for a Month approx. I saw other posts where the 1.4V/1.425V is safe to operate with. Let me know your thoughts.
 

madmatt30

Titan
Ambassador
Thanks for the Reply @madmatt30



So, should I try lowering the Load Line Calibration?



Yes, but these temps are Synthetic Burn Tests and will reach these temps with Any Video Encoding or Rendering or even the Highest CPU Intensive Games.

I did some video encoding before and after the overclock I gained a 3 Min time difference(almost) for 1.8G MP4. The Application was even using single core only. So, I think this is quite a difference in real world test.

When the CPU without any overclock works in TurboBoost Mode I found frequencies reaching 1.2900/1.3000 so thought increasing to 1.350 and test the different multiplier settings. Now its stabling at 1.35V for a Month approx. I saw other posts where the 1.4V/1.425V is safe to operate with. Let me know your thoughts.
Drop the llc a setting lower, extreme will absolutely push excess voltage both at idle and a load.
The issue here is your overclock 'may' fail with llc dropped a stage beacuse you really are at the tipping point at 4. 5ghz with an old sandy/ivy cpu.
 

CompuTronix

Intel Master
Moderator
... i5-2500k on P8Z68V at around 4.5GHZ at 1.3500V with a Hyper 212 EVO ... idles at around 41-43C ... P95 Tests at 4.5GHZ for 6 Hours ... temps of about 75-77 C Avg ... stress testing the CPU the voltages spiked to 1.360V from the base of 1.336 as well where as I have set the Voltage to 1.3500V ... Temps or Voltage shows different by different applications like GPUZ and HW Monitor ...

My Question : Why is this happening even when I set the voltage to 1.3500V? If this based on the Speedstep Option that I have enabled in the Bios which one is correct? Volts of 1.368V, Are they in safe Range and can be used for daily usage?

But when I try to Increase the Multipler from 45 to 46 or 47 keeping at the same voltage i am getting BSOD with P95 tests.

My Question : So, Should I try increasing the voltage to 1.375 and then try increasing the multiplier to 47 or 48? Will these voltages damage the CPU or motherboard? Should I not try doing that? Are these voltages in safe Range for daily usage?
... should I try lowering the Load Line Calibration? ... these temps are Synthetic Burn Tests and will reach these temps with Any Video Encoding or Rendering or even the Highest CPU Intensive Games ... stable at 1.35V for a Month approx. I saw other posts where the 1.4V/1.425V is safe to operate with.
On behalf of Tom's Moderator Team, welcome aboard!

What is your ambient (room) temperature?

Which version of Prime95 and which Torture Test?

If it's version 26.6 then that's fine, but if it's the latest version 29.8 then you need to disable all AVX test selections. Your 2500K has AVX Instruction Sets, so if Prime95 is allowed to run AVX code then your Core temperatures will be unrealistically high. The Small FFTs test is a steady-state 100% TDP workload. When run without AVX, this is how Intel validates TDP and Thermal Specifications.

GPU-Z and Hardware Monitor are not the most accurate utilities, so I would disregard those values. Instead, use Core Temp for basic monitoring and HWiNFO for very detailed monitoring. Both are frequently updated and are known to be accurate.

The Vcore you set in BIOS differs from Windows due to Load Line Calibration (LLC). When adjust LLC, only a steady-state 100% workload should be used so as to minimize voltage fluctuations.

The goal for adjusting LLC is to match as closely as possible (but not overshoot) the Vcore you set in BIOS to the Vcore you monitor in Windows during a steady-state 100% workload. Intel intends for there to be at least some small amount of "Vdroop" between BIOS and 100% workload for processor longevity. A slight undershoot is specified, but any overshoot is not recommended.

Concerning Vcore, 1.375 is about as high as you should push a 2nd generation 32 nanometer processor, such as your 2500K. No two processors are identical; each is unique in voltage tolerance, thermal behavior and overclocking potential, which is often referred to as the "silicon lottery".

Overclocking is always limited by two factors; voltage and temperature.​

Each Microarchitecture has a “Maximum Recommended Vcore”. For example, it’s important to point out that 22 nanometer 3rd and 4th Generation processors will not tolerate the higher Core voltages of other Microarchitectures.

Here's the Maximum Recommended Vcore per Microarchitecture from 14 to 65 nanometers since 2006:


We know that over time, excessive voltage and heat damages electronics, so when using manual Vcore settings in BIOS, excessive Core voltage and Core temperature can cause accelerated "Electromigration". Processors have multiple layers of hundreds of millions of microscopic nanometer scale components. Electromigration erodes fragile circuit pathways and transistor junctions which results in the degradation of overclock stability, and thus performance.

Although your initial overclock may be stable, degradation doesn't appear until later, when increasingly frequent blue-screen crashes indicate a progressive loss of stability. The more excessive the levels of voltage and heat and the longer they're sustained determines how long until transistor degradation destabilizes your overclock. Decreasing overclock and Vcore may temporarily restore stability and slow the rate of degradation. Extreme overvolting can cause degradation in minutes, but a sensible overclock remains stable for years.

Each Microarchitecture also has a "Degradation Curve". As a rule, CPUs are more susceptible to electromigration and degradation with each Die-shrink. However, the exception to the rule is Intel's 14 nanometer Microarchitecture, where advances in FinFET transistor technology have improved voltage tolerance.

Here's how the Degradation Curves correspond to Maximum Recommended Vcore for 22 nanometer 3rd and 4th Generation, which differs from 14 nanometer 5th through 10th Generation:


Degradation Curves are relative to the term “Vt (Voltage threshold) Shift” which is expressed in millivolts (mv). Users can not monitor Vt Shift. With respect to overclocking and overvolting, Vt Shift basically represents the potential for permanent loss of normal transistor performance. Excessively high Core voltage drives excessively high Power consumption and Core temperatures, all of which contribute to gradual Vt Shift over time. Core voltages that impose high Vt Shift values are not recommended.

Your 2nd generation 32 nanometer 2500K has a similar degradation curve which differs slightly from the 14 nanometer curve shown on the above graph.

To achieve the highest overclock, keep in mind that for your final 100 MHz increase, a corresponding increase in Core voltage of about 50 millivolts (0.050) is needed to maintain stability. If 65 millivolts (0.065) or more is needed for the next stable 100 MHz increase, it means you're attempting to overclock your processor beyond its capability. This also means that since you're at 4.5GHz, 4.6 might not be achievable, and anything higher would be out of reach.

Here's the nominal operating range for Core temperature:

Core temperatures above 85°C are not recommended.

Core temperatures below 80°C are ideal.


Core temperatures increase and decrease with ambient temperature, for which the International Standard for "normal" is 22°C or 72°F.

With high-end cooling you might reach your Maximum Recommended Vcore limit before you reach the ideal Core temperature limit at 80°C. With low-end cooling you’ll reach 80°C before your Vcore limit. Regardless, whichever overclocking limit you reach first is where you should stop.

Remember to keep overclocking in perspective. For example, the difference between 4.5 and 4.6 GHz is less than 2.3%, which has no noticeable impact on overall system performance. It simply isn’t worth pushing your processor beyond recommended Core voltage and Core temperature limits just to squeeze out another 100 MHz.

You might want to read this: Intel Temperature Guide

Once again, welcome aboard!

CT :sol:
 

ASK THE COMMUNITY

TRENDING THREADS