14900k high voltage

[unclewebb]

what about lite load calibration control

I do not have any experience with how best to set up a 14th Gen CPU on a MSI desktop motherboard. Use whatever setting works best. Watch for WHEA errors in HWiNFO. You want to see 0 WHEA errors or any other similar errors.

i disabled C-state as well

Did that improve game play any? Disabling the C states will also disable the 60 turbo multiplier when lightly loaded. I am not sure if you are OK with that. At default settings with the C states disabled, light loads will be limited to the 57 multiplier.

 

[ShrimpBrime]

I do not have any experience with how best to set up a 14th Gen CPU on a MSI desktop motherboard. Use whatever setting works best. Watch for WHEA errors in HWiNFO. You want to see 0 WHEA errors or any other similar errors.


Did that improve game play any? Disabling the C states will also disable the 60 turbo multiplier when lightly loaded. I am not sure if you are OK with that. At default settings with the C states disabled, light loads will be limited to the 57 multiplier.

Old friend.
Suggestion/s
1st - clear cmos or set all defaults bios
2nd - set xmp, reboot back to bios
3rd - set PL1 and PL2 limits to desired wattage (253w or lower if thermally challenged) *not auto
4th - set LLC to lvl 4 on MSI boards
5th - test settings in windows
6th adjust PL1 and PL2 settings if thermal headroom is available.
7Th - test and make appropriate LLC and PL adjustments.

Note. Whea errors are almost always feom too much v-droop or under-volting.

 

[nickdog]

so after changing to lite load mode 3 im not receiving whea errors anymore but my power is still throttling edp it says ring interconnect frequency clipped due to electrical design or other constraints maximum current consumption, SVID voltage limt or PL4 even with c-state enabled i never reached the full 6 GHZ? is it because i didnt click overclock or cpu boost in bios?

 

[unclewebb]

i never reached the full 6 GHZ

Too many Windows background tasks can interfere with the CPU using the maximum turbo multiplier. HWiNFO may not accurately report the highest multiplier when the CPU is rapidly switching between the 57 and 60 multiplier. After making any changes in the BIOS, always delete the ThrottleStop.INI file before running ThrottleStop. Try running the Cinebench single core test. Does ThrottleStop report a multiplier somewhere between 57 and 60 during a single core test?

Check that IccMax is set to the max for the CPU core, P cache, Intel GPU and iGPU Unslice. Also check that Power Limit 4 in ThrottleStop is set to the max, 1023. Any of these not being set to the max by the BIOS can cause EDP throttling.

With HWiNFO not running, does ThrottleStop Limit Reasons show any reasons for throttling lighting up red during a single core test? Can you maintain the full 57 multiplier during a Cinebench full load test? I am not sure if this is possible when your PL1 and PL2 power limits are set to the default 253W.

 

[nickdog]

Too many Windows background tasks can interfere with the CPU using the maximum turbo multiplier. HWiNFO may not accurately report the highest multiplier when the CPU is rapidly switching between the 57 and 60 multiplier. After making any changes in the BIOS, always delete the ThrottleStop.INI file before running ThrottleStop. Try running the Cinebench single core test. Does ThrottleStop report a multiplier somewhere between 57 and 60 during a single core test?

Check that IccMax is set to the max for the CPU core, P cache, Intel GPU and iGPU Unslice. Also check that Power Limit 4 in ThrottleStop is set to the max, 1023. Any of these not being set to the max by the BIOS can cause EDP throttling.

With HWiNFO not running, does ThrottleStop Limit Reasons show any reasons for throttling lighting up red during a single core test? Can you maintain the full 57 multiplier during a Cinebench full load test? I am not sure if this is possible when your PL1 and PL2 power limits are set to the default 253W.

so when i changed all the boxes to unlock voltage and set ICCmax 511.75 in every box it allowed the restriction limit seem to have vanished im still at 1.3volts though even with some undervolt setting ICC max fixed the issue with it not recieving full 6 ghz speed thanks!

nevermind it was working ok now i cant run cincbench in multi core mode anymore i didnt change anything it doesnt respond now please help im seeing cpu error again and ring limit reason again

 

[unclewebb]

I suggested setting Power Limit 4 to a value of 1023. It looks like you did not do that. Maybe that will help.

i cant run cincbench in multi core mode anymore

What happens? Does Cinebench immediately crash or freeze? If your LLC setting is reducing the CPU voltage then undervolting on top of that is probably not a good idea. Try setting your undervolt settings to zero.

Your ThrottleStop screenshot shows that at least one core was running pretty damn close to 6 GHz. At default settings without overclocking, only the two fastest cores are capable of running at 6 GHz. The rest of the P cores are limited to 5.7 GHz.

im still at 1.3volts

That might not be enough to reliably run your CPU at its full rated speed.

 

[nickdog]

I suggested setting Power Limit 4 to a value of 1023. It looks like you did not do that. Maybe that will help.


What happens? Does Cinebench immediately crash or freeze? If your LLC setting is reducing the CPU voltage then undervolting on top of that is probably not a good idea. Try setting your undervolt settings to zero.

Your ThrottleStop screenshot shows that at least one core was running pretty damn close to 6 GHz. At default settings without overclocking, only the two fastest cores are capable of running at 6 GHz. The rest of the P cores are limited to 5.7 GHz.


That might not be enough to reliably run your CPU at its full rated speed.

i have fixed the ring throttling i believe i had to set pl1 pl2 and pl4 all to 500 and hit lock on pl 4 i passed cinebench again with 0 errors but i reduced clock speed a litle bit so my rig doesnt burn oh yeah another thing i did was change the LLC from (auto) to load line mode 3 im using lite mode 3 i think

 

[unclewebb]

The PL1 and PL2 MSR (static) power limits are still set to 253W. How did you set these to 500? The BIOS might not be working correctly.

0 errors running Cinebench sounds good. It takes a while and lots of testing to get all of the various settings just right. You are making good progress.

Intel CPUs run red hot when fully loaded running Cinebench if you try to run them at full speed. It is what it is. Buy a better cooler, slow your CPU down or just ignore temperatures in the high 90°C range. Intel says their CPUs can handle those kind of temperatures.

Edit - If you used ThrottleStop to change the power limits, you have to exit and restart HWiNFO before it updates these changes. HWiNFO does not update the static power limit information that it reports in real time.

 

[nickdog]

The PL1 and PL2 MSR (static) power limits are still set to 253W. How did you set these to 500? The BIOS might not be working correctly.

0 errors running Cinebench sounds good. It takes a while and lots of testing to get all of the various settings just right. You are making good progress.

Intel CPUs run red hot when fully loaded running Cinebench if you try to run them at full speed. It is what it is. Buy a better cooler, slow your CPU down or just ignore temperatures in the high 90°C range. Intel says their CPUs can handle those kind of temperatures.

Edit - If you used ThrottleStop to change the power limits, you have to exit and restart HWiNFO before it updates these changes. HWiNFO does not update the static power limit information that it reports in real time.

i set them through throttlestop how about now it looks better i changed it in bios but i still need throttlestop running and max iccmax on all options ill get RING:max Vr voltage,ICCMAX,PL4 limit reason in ring limit but it dissapereas after i make sure all PL are set in throttlestop seems better right? this time i got 1982 instead of 1928 benchmark score do you think selecting Msi ultimate performance for cpu cooler tuning 4096w will fix it or could that ruin my stuff i didnt see no limits in throttle stop anymore with the things you told me to adjust looks good so far also is it smart to have TVB disabled

 

[BoggledBeagle]

Do not use any special software. All you need is to limit core frequencies and power limits in BIOS. My 14900K limited to 5400 MHz does not go above 1,3V.

 

[nickdog]

Do not use any special software. All you need is to limit core frequencies and power limits in BIOS. My 14900K limited to 5400 MHz does not go above 1,3V.

how do i limit them to 5400 Mhz wihthout throttlestop do i go to active per core and enter the number i see a thing that says per p core auto offset and other setting do i manually have to put in 54 instead of 57

 

[BoggledBeagle]

Yes, you type 54x or whatever for each P core.

BTW you can limit E cores a bit too, they are very densely packed and the heat stress they create per silicon area is high too.

 

[Outback Bronze]

Do not use any special software. All you need is to limit core frequencies and power limits in BIOS. My 14900K limited to 5400 MHz does not go above 1,3V.

This is exactly what I have been doing to Raptor Lake for years with my P & E cores. I also then apply a negative offset voltage of around -0.100 give or take for an even lower voltage which in turn gives me better thermals and power consumption.

 

[BoggledBeagle]

Negative voltage offset is not good in the situation of potential degradation. The "excess" voltage is there to ensure stability. Throwing away 2/3 or more of your safety voltage buffer is dangerous, because as the chip naturally slowly degrades (every chip does this), it may lose stabilty in some not so obvious way and for example corrupt data without you noticing that before it is too late.

 

[Outback Bronze]

Negative voltage offset is not good in the situation of potential degradation. The "excess" voltage is there to ensure stability. Throwing away 2/3 or more of your safety voltage buffer is dangerous, because as the chip naturally slowly degrades (every chip does this), it may lose stabilty in some not so obvious way and for example corrupt data without you noticing that before it is too late.

Yep, ok fair call I just thought I would be prolonging the processors life even more with reduced voltage feed than it needed to be. A negative offset can always be readjusted.

I haven't had any issues on this particular 14900k for the past (almost) 2 years with a negative offset, even under heavy loads or idle.

 

[BoggledBeagle]

Yes, if this was obvious and easilly measurable and observable, you could be safely doing it. For example you had an initial 150 mV safety margin, you removed 100 mV from it, and then you somehow knew that the chip after a year degdaded by for example 30 mV, you could lower your neg. offset to 70mV and still be safe.

 

[unclewebb]

4096w

The previous screenshot you posted shows your CPU is within 4°C of thermal throttling when power consumption is at 207 Watts.

Leaving the turbo power limits set to the 253W default value should not be causing any significant power limit throttling. It might throttle for a millisecond or two which could trigger a yellow warning box to light up in Limit Reasons. Practically speaking, you will probably never notice any performance loss with the power limits at their default 253W setting.

You can choose to set both power limits to 500W or the Unlimited value, 4096W. It will not likely hurt anything but it will also not significantly help anything. Thermal throttling is more likely to become an issue for you before power limit throttling does. It might be your Power Limit 4 setting that was causing the EDP throttling.

Do not use any special software.

ThrottleStop has access to some settings within the CPU like the Lock settings that most BIOS versions do not give you access to. ThrottleStop might also let you set some current limits that the BIOS does not let you set.

Anyone running Windows Defender should always be running ThrottleStop. Defender can randomly slow down 13th and 14th Gen CPUs just like it was doing to 10th Gen CPUs when this problem was first discovered.

Windows Defender can Significantly Impact Intel CPU Performance, We have the Fix

Kevin Glynn, aka "Uncle Webb," our associate software author behind popular utilities such as ThrottleStop and RealTemp, developed a new utility named Counter Control, which lets you monitor and log the performance counters of Intel Core processors since 2008 (Core "Nehalem"). During development...

www.techpowerup.com

The most recent problem is some specific games are lowering the Per Core turbo limits which can significantly lower the performance of recent Intel CPUs. ThrottleStop has a fix for this issue. There is no setting in any BIOS that can fix this problem.

S

Thread 'Multiplier limited, trying to achieve max performance'

Jul 19, 2025

Good evening, I have been using Throttlestop (9.6) with basic for several years, and I am very happy with the application overall. I recently noticed a strange issue where when playing certain games (Dune awakening) my CPU was throttling to about 3500mhz. Opposed to playing World of Warcraft, where I would usually throttle to about 4400mhz. According to the logs and settings, I was only experiencing an "EDP Other" limit in the limits window.

In the course of researching possible solutions, I noticed Throttlestop 9.7 had been released, so I deleted my ini file and installed the new...

• SeawolfX13
• Replies: 4
• Forum: ThrottleStop

how do i limit them to 5400 MHz

Before limiting anything ask yourself, what problem are you trying to solve? Your CPU is designed to play games all day and all night at 5700 MHz. Intel designed their CPUs so they can run reliably even when they are hot, hot, hot. They are capable of managing themselves without any user intervention. You are not going to hurt anything by using your CPU at its Intel rated speed. If you do ever hurt it, the 14th Gen CPUs have an extended 5 year warranty so you are well covered.

 

[BoggledBeagle]

Before limiting anything ask yourself, what problem are you trying to solve? Your CPU is designed to play games all day and all night at 5700 MHz. Intel designed their CPUs so they can run reliably even when they are hot, hot, hot. They are capable of managing themselves without any user intervention. You are not going to hurt anything by using your CPU at its Intel rated speed. If you do ever hurt it, the 14th Gen CPUs have an extended 5 year warranty so you are well covered.

Are you serious?

Have you not noticed the huge degradation scandal, with many CPUs running without problem for just a few months, before they degraded?

Do you really think that people should kill several CPUs by running them at setting that kill them in a period of 5 years, or just run one CPU in a way it can handle? It would be not only a cost for Intel, but also creating waste completely needlessly. And what about the annoying hassle of RMA-ing, sending out, and swapping all those CPUs?

 

[ShrimpBrime]

Before limiting anything ask yourself, what problem are you trying to solve? Your CPU is designed to play games all day and all night at 5700 MHz. Intel designed their CPUs so they can run reliably even when they are hot, hot, hot. They are capable of managing themselves without any user intervention. You are not going to hurt anything by using your CPU at its Intel rated speed. If you do ever hurt it, the 14th Gen CPUs have an extended 5 year warranty so you are well covered.

100% It will just throttle effective clocks. User simply just needs adequate cooling.

The IA EDC throttling is just a result of the under-volt restriction. Its below (v-core) the standard threshold, so its just reported as such. I believe that was a concern of OP, but lack of comprehension on their part simply just brings the concern. Theres actually nothing wrong, the under-volt is working.

Since we all approach it differently, I have used TS in the same fashion as XTU. It simply looks different, and maybe some acronyms are shorter longer or different. I feel some of these need deep explanation especially when using HWInfo64. Just because its reporting something, doesnt mean its bad.

 

[nickdog]

so i seem to be stable at this setting lite load mode 2 LLC mode 3 limited to 5.4ghz and undervoltt adaptive offset 0.056 never reached over 1.2volts playing games also put PL back to 300w works fine but with max icc i think i found an issue with the gpu limiting power instead of cpu now im getting gpu performance limit reliability voltage everytime i run game it reaches 100% volt limit i think thats whats causing the fps drop and performance cut in half no idea what to do here sucks i dont want to mess things up but i want to make sure my pc getting max performance i already tried maximum performance mode in nvidia control panel i cant even change the power limit more then 100% on msi afterburner its a gpu which saves power and fans stop until gpu hits certain threshold