• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Getting Less MultiCore Ghz on i5 13400F

Joined
Aug 9, 2024
Messages
58 (0.17/day)
Processor i5 13400F
Motherboard Asus Z790 Prime D4
Cooling DeepCool 120MM
Memory 32GB Corsair 3200Mhz CL16
Video Card(s) Gigabyte RTX 3070
Storage Samsung 980 Pro
Display(s) LG UltraGear 24 165hz
Case DeepCool
Power Supply CoolerMaster 750W Gold
Mouse Logitech G305
Before Updating my Bios to the latest version, the microcode X129 from intel. my i5 13400F gave me 3.88Ghz on 100% Usage without the Drop.
but since the update, I only get this speed for just like 10 to 15 sec and then it drops to 3.5Ghz or 3.2Ghz.
I checked my CPU power with AIDA64 and it goes from 89W to 35W when it clockes down. ( should I put the Voltages too? )
I have a 240MM Cooler but my CPU is around 55 to 68c on full load so I'm not terminal-limited.
I'm just worried that my CPU is cooked or not. I have tested my CPU with Benchmate and it passes just fine.
 
The Bios update may have messed/changed your Power Limit settings, especially possible if your new Bios has the microcode fix for 13/14 gen processors although I don't believe processors below 13/14600K's were affected.
 
but since the update, I only get this speed for just like 10 to 15 sec and then it drops to 3.5Ghz or 3.2Ghz.

Below you can see the Intel 12700KF frequencies per thread on OpenBSD.

IlogXXv.png


The fact that your CPU clocks back to lower frequencies after a few seconds is normal.
My BIOS already had this setting that it switches to lower frequencies after a specific number of seconds.
 
The Bios update may have messed/changed your Power Limit settings, especially possible if your new Bios has the microcode fix for 13/14 gen processors although I don't believe processors below 13/14600K's were affected.
Hi, I'm using the default setting in bios.
are u sure because they say all of the TDP 65 and above are effected
 
13400f is still alder lake architecture rebranded as 13th gen. I say remove all power limits and set the voltage to fixed 1.25v-1.3v or whatever the chip needs to pass as stable.

The Bios update may have messed/changed your Power Limit settings, especially possible if your new Bios has the microcode fix for 13/14 gen processors although I don't believe processors below 13/14600K's were affected.
or maybe it's the OS, aren't some microcodes pushed through windows updates ? I'd check power settings in control panel too.
 
13400f is still alder lake architecture rebranded as 13th gen. I say remove all power limits and set the voltage to fixed 1.25v-1.3v or whatever the chip needs to pass as stable.


or maybe it's the OS, aren't some microcodes pushed through windows updates ? I'd check power settings in control panel too.
What do you mean? you mean the i5 13600K and less are based on the Alder Lake and not effected by this problem?
how about i5 13500?


can you explain it more plz.♥
 
Before Updating my Bios to the latest version, the microcode X129 from intel. my i5 13400F gave me 3.88Ghz on 100% Usage without the Drop.
but since the update, I only get this speed for just like 10 to 15 sec and then it drops to 3.5Ghz or 3.2Ghz.
I checked my CPU power with AIDA64 and it goes from 89W to 35W when it clockes down. ( should I put the Voltages too? )
I have a 240MM Cooler but my CPU is around 55 to 68c on full load so I'm not terminal-limited.
I'm just worried that my CPU is cooked or not. I have tested my CPU with Benchmate and it passes just fine.
Asking if you clear RTC after the bios flash or did you just post straight to windows.....

Set windows power plan to defaults as you read this....

Shut down the PC fully, turn off PSU.
Using a jumper, Clear RTC (cmos jumpers) 30 seconds.
Remove the jumper. Turn on PSU.
Start bios, go to settings, set all defaults.
F10 Restart
Enter Bios, configure your memory only (XMP/DOCP enabled only). Do not touch any Cpu settings otherwise.
F10 restart Enter windows and test again.
If you change windows power plans, I suggest a restart of the PC after the change is made.

My thoughts. 2 bits a gander.
GL
 
Last edited:
The Bios update to the microcode doesn't look for what Processor you have, it simply alters the power limits for all Intel CPUs and sets them to the default Intel spec.
 
1.25 MB L2 Cahce is Alder lake die even if it shows Raptor lake in CPUz that kind of thing.
Set the Short duration power limit to 148 in Watts same as the long duration.
Leave all voltages alone since 4.1GHz all P-core needs even less than default. No point in going any higher that would raise the power.

And ironically Intel did just that with the patch, they cut off the power limit as they raised the voltage across the board to make sure the already supposedly degraded overclocking chips were working properly. And they broke the chips that weren't broken otherwise, now they can't reach their initial speeds and performance is degraded and they enter a world of pain. Planned obsolescence at its best.
 
What do you mean? you mean the i5 13600K and less are based on the Alder Lake and not effected by this problem?
how about i5 13500?


can you explain it more plz.♥
I think this explains it better than I could, this is a news piece but later actual reviews of the processor show it's pretty accurate ................

Non-K 13th Gen Core i5 (such as i5-13400) Based on Older "Alder Lake" Architecture, Hints Intel Slide | TechPowerUp

Review after news piece above of a 13500 highlighting the lower L2 cache levels of Alder Lake .............

Intel Core i5-13500 Review | TechSpot
 
I think this explains it better than I could, this is a news piece but later actual reviews of the processor show it's pretty accurate ................

Non-K 13th Gen Core i5 (such as i5-13400) Based on Older "Alder Lake" Architecture, Hints Intel Slide | TechPowerUp

Review after news piece above of a 13500 highlighting the lower L2 cache levels of Alder Lake .............

Intel Core i5-13500 Review | TechSpot
Hi, thanks so what does this mean is that my CPU is Safe?
but why does it do this?
after Bios Update my temps are better than before but I didn't have problem even before the Update. My CPU was around 70c at full load now it's on 63c. but in a cost on going down from 3.88Ghz to 3.54Ghz or something around this.
 
I agree with ShrimpBrime - it is probably a good idea to reset the BIOS back to the defaults before updating the firmware. If you have made several changes, you might want to record them first.

Asking if you clear RTC after the bios flash or did you just post straight to windows.....

Clear RTC (cmos jumpers)
Right idea, not so right terminology.

Resetting the BIOS deletes or "clears" the CMOS data from the CMOS device's memory and it "resets" the RTC (real-time clock) back to zero.

The CMOS data contains the user changes to the BIOS firmware default settings. A prime example of this information is the "boot order", telling the system which drive to look for first for an OS.

It also includes the system date and time, but note the date and time are NOT the same as the RTC (real-time clock) information.

The RTC may be a separate device or integrated in the CMOS device. Either way, the RTC is a discrete device by function, even if physically integrated with the CMOS device. The RTC is simply a counter that counts the cycles of an oscillator - commonly running at 32.768KHz. That frequency is NOT arbitrary but specifically used because in binary circuits, it equals 2^15 (2 to the power of 15) and provides a time period of precisely 1.0 second with a 15 stage binary counter.

So the RTC does not know what the date and time are, it only knows how many oscillations (thus seconds) have elapsed since it was reset back to zero.

When you set the date and time in the BIOS, you are just resetting the RTC again, only this time, not to zero but to a value (number of oscillations) that "represents" the current date and time.

Clear as mud, huh?

The CMOS battery is there not just to keep the data containing the user changes to the BIOS firmware "alive" but also to keep the RTC oscillator oscillating while the computer is turned off (and the PSU unplugged or its master power switch set to off).

Side note. "CMOS" is actually a misnomer these days. Originally, more than 60 years ago, CMOS was an discrete (individual and separate) electronics component, a type of memory device and stood for "complementary metal–oxide–semiconductor". Fast forward 40 - 50 years and CMOS these days on motherboards refers to an entire circuit consisting of several components. Either way, the concept remains the same.

Note too that CMOS memory devices were chosen for this purpose because they are easy for the user to reset (delete all stored data). All that's needed to totally erase CMOS data is to remove the "holding" voltage from the device. Then almost instantly (within just a few clock cycles) the data is gone. Had motherboard engineers and designers wanted the CMOS data to be hard to delete/reset, they could have easily used a different device, like an EEPROM instead.

The fact that your CPU clocks back to lower frequencies after a few seconds is normal.
I agree with this. When the demand on the CPU decreases, there is no reason for it to keep running at full speed. However, when the demand on the CPU increases, the frequency should too - up to the CPU's maximum capability.

but in a cost on going down from 3.88Ghz to 3.54Ghz

Are you saying 3.54GHz is as high as it goes, even when stressed?

I must confess, I am a bit confused because according to the Intel ARK for the i5 13400F, the maximum single core freq should be 4.6GHz. :confused: I note according to Intel, the Microcode (0x129) patch does not prevent users from overclocking so it seems to me it should not prevent the CPU from throttling up to its published "turbo" max speed of 4.6GHz.

I note on this ASUS page, it says the latest BIOS version is a "Beta" version. Unfortunately, there's no way to really tell what to read into that.
 
Hi, thanks so what does this mean is that my CPU is Safe?
but why does it do this?
after Bios Update my temps are better than before but I didn't have problem even before the Update. My CPU was around 70c at full load now it's on 63c. but in a cost on going down from 3.88Ghz to 3.54Ghz or something around this.
You have said the Power Limit settings are "default" , your new Bios may have changed the default, what exactly does it say for PL1 and PL2?
 
Every review is a pain in the ass to read. Everyone compared the chip to a bunch of other chips, so the data for a 13400F is all mixed in with all the other specs of various processors.

Floating point all core load clocks at 148w should reach between 3.9ghz and 4.1ghz. This might depend on the AVX load as well. I'm assuming 4.1ghz is with power lifted. Can't tell ya, cause I haven't one to test with unfortunately.

The 3.88ghz reading is simply from the bus droop which is an EPA compliance thing. Some boards will allow the user to fix BCLK to 100mhz. Look in your bios for this. My MSI board has it located in the cpu configuration menu where you enable and disable cores for example.

@Bill_Bright- thanks for that explanation of CMOs vs RTC. On modern boards, the cmos jumper is also the rtc jumper.

In bios you can clear cmos settings without using the cmos (RTC) jumper. I always suggest to use the jumper anyways, it clears everything.

Time in bios can and will be updated through UEFI and windows. Just have to set the time date accordingly, but can change it either in bios or windows, just depends on what the user wants to so..

But again, thank you!

GL OP!
 
On modern boards, the cmos jumper is also the rtc jumper.
Not always. You can tell if the data and time stay correct after resetting the CMOS.

I too use the jumper. Or you can remove the CMOS battery for a few seconds. The jumper simply shorts the battery voltage to ground, thus removing the voltage from the CMOS device. Same effect as removing the battery. This is why, when you see folks recommend removing the battery and moving the jumper, they really don't know what they are talking about. lol

I note some boards now come with a reset switch. Those are really handy, and just as effective as using the jumper or removing the battery. However, they are much safer too. They remove the potential ESD damage to sensitive devices. This is because you typically don't have to reach in with a hand potentially charged with 10s of 1000s of volts of static electricity that's itching to jump to ground through a sensitive device.
 
Every review is a pain in the ass to read. Everyone compared the chip to a bunch of other chips, so the data for a 13400F is all mixed in with all the other specs of various processors.

Floating point all core load clocks at 148w should reach between 3.9ghz and 4.1ghz. This might depend on the AVX load as well. I'm assuming 4.1ghz is with power lifted. Can't tell ya, cause I haven't one to test with unfortunately.

The 3.88ghz reading is simply from the bus droop which is an EPA compliance thing. Some boards will allow the user to fix BCLK to 100mhz. Look in your bios for this. My MSI board has it located in the cpu configuration menu where you enable and disable cores for example.

@Bill_Bright- thanks for that explanation of CMOs vs RTC. On modern boards, the cmos jumper is also the rtc jumper.

In bios you can clear cmos settings without using the cmos (RTC) jumper. I always suggest to use the jumper anyways, it clears everything.

Time in bios can and will be updated through UEFI and windows. Just have to set the time date accordingly, but can change it either in bios or windows, just depends on what the user wants to so..

But again, thank you!

GL OP!
Hi, I'm using my PC for Video editing and I usually Render My files with Nvenc with Handbrake.
before the Update when I used handbrake or any other app that makes my CPU Work Like Premiere rendering and .... for 100% Usage, My CPU Was on 3.88Ghz On TaskManager. it's normal because of the mix of P and E cores Clock speed. my P cores were on 4.1Ghz with no problem but since the Update, my CPU on full load only gives me the 4.1Ghz or in another word ( 3.88Ghz ) for just a few seconds and then it drops to 3.5Ghz or 3.3Ghz ( or 3.8Ghz on P Cores instead of 4.1Ghz ).

i just changed my Bios Option from Intel Default to Asus and remove the Limitation On 90c. and my CPU is working fine.

the only thing that bothers me is that from my understanding the 13500, 13600, 13400 are based on the older architecture and not effected by the intel recent problem right? so I can freely Work With my CPU to get the Full out of it?
 
Not always. You can tell if the data and time stay correct after resetting the CMOS.

I too use the jumper. Or you can remove the CMOS battery for a few seconds. The jumper simply shorts the battery voltage to ground, thus removing the voltage from the CMOS device. Same effect as removing the battery. This is why, when you see folks recommend removing the battery and moving the jumper, they really don't know what they are talking about. lol

I note some boards now come with a reset switch. Those are really handy, and just as effective as using the jumper or removing the battery. However, they are much safer too. They remove the potential ESD damage to sensitive devices. This is because you typically don't have to reach in with a hand potentially charged with 10s of 1000s of volts of static electricity that's itching to jump to ground through a sensitive device.
I like boards with clear cmos on the I/O panel. That is a very smart implementation for sure. Not only for ease of access, but as you mentioned the risk of ESD.

I've many times used an old case reset switch and wire it to the cmos jumpers. Mainly to make it easier to get at. I have a couple itx matx boards the jumper is under the video card. Gotta pull the card out to get at it. What a pain in the butt.
 
With this new microcode you are likely getting 50 mV higher voltages.


as shown by tech yes, jayz2c didn't so my guess is that it clocks down to stay within 88 Watts but for the lack of better management it drops too much to 3.5 Ghz instead of doing the more precise thing of 4.0 then 3.9 and so on.
 
I've many times used an old case reset switch and wire it to the cmos jumpers. Mainly to make it easier to get at.
I keep a couple of these on my bench, and one in my tool bag I take on house calls that work for that purpose too.
 
I keep a couple of these on my bench, and one in my tool bag I take on house calls that work for that purpose too.
Haha, perfect. Good stuff.

Hi, I'm using my PC for Video editing and I usually Render My files with Nvenc with Handbrake.
before the Update when I used handbrake or any other app that makes my CPU Work Like Premiere rendering and .... for 100% Usage, My CPU Was on 3.88Ghz On TaskManager. it's normal because of the mix of P and E cores Clock speed. my P cores were on 4.1Ghz with no problem but since the Update, my CPU on full load only gives me the 4.1Ghz or in another word ( 3.88Ghz ) for just a few seconds and then it drops to 3.5Ghz or 3.3Ghz ( or 3.8Ghz on P Cores instead of 4.1Ghz ).

i just changed my Bios Option from Intel Default to Asus and remove the Limitation On 90c. and my CPU is working fine.

the only thing that bothers me is that from my understanding the 13500, 13600, 13400 are based on the older architecture and not effected by the intel recent problem right? so I can freely Work With my CPU to get the Full out of it?
Yes, the cpu should be unaffected by the supposed instability issues. Just continue using it normally and you'll be fine.

The core P is Alder Lake C0 stepping. So that's how we determine if the cpu would be effected. And you should have no issues.

Looks like from what you are saying is the boost duration has been limited. So the cpu will just boost for less time and settle in to 3.5ghz all core.
 
Looks like from what you are saying is the boost duration has been limited. So the cpu will just boost for less time and settle in to 3.5ghz all core.
A very strange thing for you to say, for a socket 1700 owner that is. That the duration is the problem when the power is the problem. Intel went out of their way to do two things to make sure crashes don't happen by raising the voltage by 50 mV for all core but by doing so the power is now slightly elevated. and lowering the voltage by 50 mV for 1-2 core boosting so that it is limited to 1.55. And by what sounds like it broke the whole power management and ecosystem even. It need more fine tuning.
 
A very strange thing for you to say, for a socket 1700 owner that is. That the duration is the problem when the power is the problem. Intel went out of their way to do two things to make sure crashes don't happen by raising the voltage by 50 mV for all core but by doing so the power is now slightly elevated. and lowering the voltage by 50 mV for 1-2 core boosting so that it is limited to 1.55. And by what sounds like it broke the whole power management and ecosystem even. It need more fine tuning.
This duration is the problem for everyone except myself actually. Same thing with power and all that.

Because it seems everyone else would like Intel and Windows to manage their cpu.

I on the other hand am not letting the system control my processor. Almost all my clocking on Z series boards are controlled by myself, the user.

And thus with my applicable cooling, 5.8ghz P-Cores and 4.7ghz E-Core with phenomenal performance runs to my liking.

Everyone has some undaunting fear of degradation. Sure, this has been a topic about processors since single and dual core days some 2 decades ago or more. You'd figure by now, these click bait title watch my you tube video make me rich sites like LTT and GN would be a little more forthcoming with truth.

Again. It's extremely difficult to break a processor. Everyone gets hung up on v-core.

Shit man. 1.55v isn't a big deal for running a single core, or even 2 cores. You're just not pulling that many amps. Maybe 60 to 80w to run a couple cores at 1.55v. Not a big deal.

But hey, click that bait. have seen TONS of threads about degradation. And (almost?) ZERO threads of end users (50% of users as click bait suggests) going to forums or social media actually having a bad cpu.

Take that as you want. Can pick my brain all you want, I'm open to telling my opinion. I do not require people to believe my thoughts either.
 
The maximum power setting of 4095 is basically unlimited. The max time window is 448 seconds and then it ends. Just undervolt the CPU by -100mV and see if it holds 4.1 in this case. Just in case reset the bios and the windows power plan.
 
The maximum power setting of 4095 is basically unlimited. The max time window is 448 seconds and then it ends. Just undervolt the CPU by -100mV and see if it holds 4.1 in this case. Just in case reset the bios and the windows power plan.
I'd just go in and apply the max multiplier, set XMP and let it fly.

Under-volt? It runs hot? Put a bigger cooler on the 13400F, It'll run cooler and quieter this way. There should be absolutely no reason to under-volt this cpu.
 
Back
Top