Question i9-10900kf not running properly?

Feb 24, 2021
9
1
15
0
So i decided to get a new pc a couple weeks ago and decided to get a decent one. I was running some tests today and noticed my cpu was not running how it is supposed to. Im assuming my brand new 10900kf is not supposed to be slower than a 10850k :D. The pc is a lenovo prebuilt could not get a 3080 anywhere so i decided to go for this since i could get it fast. rest of the specs are msi rtx 3080, Motherboard Lenovo's 3715, Ram no clue what brand but 16gigs at 3200 dual channel. All the drivers are up to date and im out of ideas what to do so i come crawling to you guys in the time of great need give me some ideas what could be wrong and what i should do?
 
You can certainly bet that any Lenovo is likely not running XMP-ram at 3600 MHz, nor is MCE-enabled, and, likely enforces standard clock speed boost limit/duration protocols, and, at least attempts to abide by the Intel-defined 125 watt TDP...

Someone else's enthusiast-configured 10850K-based system on a good Z490 mainboard running 4000+ MHz RAM profiles, configured with great cooling, with all the aforementioned clock speed/boost duration/power limits removed will almost certainly be ahead by 10-15% in a large variety of benchmarks...or more.
 

Mr.Spock

Respectable
Dec 8, 2019
1,110
130
1,890
107
wouldn't rely on the comparison to reference as your sole gauge of speed. have seen OC'ed processors show less than reference in Single thread, probably due to differences in driver and background jobs.
 
Feb 24, 2021
9
1
15
0
You can certainly bet that any Lenovo is likely not running XMP-ram at 3600 MHz, nor is MCE-enabled, and, likely enforces standard clock speed boost limit/duration protocols, and, at least attempts to abide by the Intel-defined 125 watt TDP...

Someone else's enthusiast-configured 10850K-based system on a good Z490 mainboard running 4000+ MHz RAM profiles, configured with great cooling, with all the aforementioned clock speed/boost duration/power limits removed will almost certainly be ahead by 10-15% in a large variety of benchmarks...or more.
Ok maybe 10850k was not the best example to use my bad. But lets say i9-9900kf even that beats it currently
 

Phaaze88

Polypheme
Ambassador
Hmm...
1)Cpu isn't turbo boosting.
2)Gpu looks ok; it still landed within 10% of the average, which I would consider normal for stock settings.[Its performance is easily affected by thermals and power settings though.]
3)Those Kingston A400 drives don't hold up very well. Performance drops off more than usual as it fills up, of which it's more than 70% full already.[I keep forgetting if it's 70 or 80% when storage drives normally start slowing down...]

Number 1 is the biggest flag by a long shot. That's something either Lenovo did in the bios, or something else is going on...
Did you update all the drivers from Lenovo's product support website?
How hot does the cpu run under loads?
 
Feb 24, 2021
9
1
15
0
Hmm...
1)Cpu isn't turbo boosting.
2)Gpu looks ok; it still landed within 10% of the average, which I would consider normal for stock settings.[Its performance is easily affected by thermals and power settings though.]
3)Those Kingston A400 drives don't hold up very well. Performance drops off more than usual as it fills up, of which it's more than 70% full already.[I keep forgetting if it's 70 or 80% when storage drives normally start slowing down...]

Number 1 is the biggest flag by a long shot. That's something either Lenovo did in the bios, or something else is going on...
Did you update all the drivers from Lenovo's product support website?
How hot does the cpu run under loads?
I did some further testing and for some reason userbenchmark does not want to read the boost clocks? i ran another test with userbenchmark and i had HWmonitor on wich showed the cpu boosting to over 4.1 and for some reason userbenchmark does not record it lol? Im assuming this is becaus of lenovo's trash mobo. Cpu under load around 50ish
 
Install/run both HWMonitor and CPU-Z.

In CPU-Z, select the 'bench' tab, and click on 'stress CPU', noting clock speeds achieved in HWmonitor...

It would be nice to see sustained 4.7 GHz on all cores, but, if manufacturer decides to strictly adhere to Intel specs, it's possible you will see the clock speeds lower themselves to something above 3.7 GHz, but less than 4.7 GHz, after 30-120 seconds, which will vary depending on BIOS implementation/settings/manufacturer lust for performance/temp considerations.

As your results would be compared to many other folks with Z-series boards and OC's, do not really expect to keep up with OC'd 9900K, 10850K, in the land of dubious Userbench, etc., etc...

(You might also see what default behavior might be altered within Intel's XTU, perhaps it will allow override of boost duration and/or power limits; can't hurt!)
 
Last edited:
Feb 24, 2021
9
1
15
0
Install/run both HWMonitor and CPU-Z.

In CPU-Z, select the 'bench' tab, and click on 'stress CPU', noting clock speeds achieved in HWmonitor...

It would be nice to see sustained 4.7 GHz on all cores, but, if manufacturer decides to strictly adhere to Intel specs, it's possible you will see the clock speeds lower themselves to something above 3.7 GHz, but less than 4.7 GHz, after 30-120 seconds, which will vary depending on BIOS implementation/settings/manufacturer lust for performance/temp considerations.

As your results would be compared to many other folks with Z-series boards and OC's, do not really expect to keep up with OC'd 9900K, 10850K, in the land of dubious Userbench, etc., etc...

(You might also see what default behavior might be altered within Intel's XTU, perhaps it will allow override of boost duration and/or power limits; can't hurt!)
So i ran cpu-z cpu stress and as soon as i hit stress all the clock went down from around 4.1ish to 3.5 and would not go above 3.6
 

ASK THE COMMUNITY

TRENDING THREADS