30452: dGPU is hotter/clock is higher since I upgraded to Windows 10 19H1 v1903
 

Im having this issue with WUDFLoad and windows 10 on the latest (and previous version of windows 10) currently running 1803 and the latest Display Link drivers

It seems anything that requires drawing items on the external displays, causes wudfhost to spike (seen noticeable if you move the mouse around), WUDFHost is consistantly at the top of my cpu / average cpu load in resource monitor.

Has anyone found an actual solution to this.

Im running a Dell XPS 15 with GTX 1050, with a Dell Dock D3100 USB 3 dock.

Ive noticed performance issues for a while now, but only recently have i decided to really dig into what it could be. So im not sure exactly when it started.

same issues - brand new Lenovo E490 with Lenovo Hybrid C Dock
 

Wow, I never expected so many people out there facing the same issue.

So, bought a brand new Lenovo E490 and a Lenovo Hybrid-C Dockingstation. Connected two external monitors via HDMI and the fan party began. Even my old L440 never was that loud.

Each and every single driver and BIOS is up to date, but as soon as I connect one or two monitors via docking station, the CPU temperature increases by 10 degrees. This is just not acceptable!

Just did a simple testing-setup: Fullscreen Youtube video, thats it.

Two monitors via HDMI on Dock: CPU constantly between 65 and 70 degrees
One monitor via HDMI on Dock: CPU constantly between 65 and 70 degrees
One monitor via HDMI directly on Notebook: CPU constantly between 55 and 60 degrees.

Come on displaylink, are you guys for real?

dGPU is hotter/clock is higher since I upgraded to Windows 10 19H1 v1903
 

Same issue with windows 10 after the displaylink software update, additionally i fixed some processes of WUDFHost.exe and windows modules installer worker responsible for higher usage but the problem remains.

See my attached Screenshots for results:

"zweiMonitoreAnDock" = two monitors via dockingstation Lenovo Hybrid-C
"einMonitorAnDock" = one monitor via dockingstations
"einMonitorAnNotebook" = one monitor directly plugged into my lenovo E490

All I do is playing a fullHD youtube video. I waited for five minutes after each switch so that temperature could settle.

Two monitors: ~67°
One monitor via dockingstation: ~65°
One monitor directly plugged into notebook: ~55°

Is this the way it's supposed to work?

So, got myself another Lenovo dock, this time the USB-C dock, for testing purposes. It only allows 2 x FullHD via VGA and displayport, but for testing that's fine.

Result: It works like a charme. While running the exact same programs as with the Hybrid-C docking station, temperature is around 50° instead of over 60°.

So, displaylink people: Why does a brand new Hybrid-C lenovo with each and every driver up to date heat up my CPU by over 10° compared to a lower-end USB-C lenovo dock?

Why is it the users who have to keep testing and do the job you guys should do?

dGPU is hotter/clock is higher since I upgraded to Windows 10 19H1 v1903
 

Hi Guys and girls,

Ive done some scouting and Ive found myself here.

I have a Acer Predator H300 with a i7700hq / GTX1060, tweaked for cool temps to allow as much use as possible out of the hardware including a liquid metal application. So belive me, without the dock, its VERY cool.

Bought one of these docks as i need t power multiple monitors (mostly just one external).

The dock is used in an Autocad Enviroment 90% of the time with some light browsing and sometimes gaming.

So Today I formatted 3 times to be completely sure of my result.

So, laptop formatted no display link, functioning temperatures normal and very low.

I installed the station and drivers, at first temps normal but as I start to work or do anything temps start to rise to a point where if im saving an Autocad Drawing thats heavier with more xrefs due to the dock it spikes and rises to 75-80 deg. This is not normal. I tested this in another scenario without the dock and the temps dont even compare.

I installed and tried every driver available, no dice. I also reformatted 3 times to try and replicate this issue, to make sure it was nothing else, Like I said, laptop is undervolted and tweaked and lm applied.


im running the latest build of 1903. Now. in the past with 1703/1709 I did not have any issues even say up to 1809 I did not have issues, temps were always decent while working with it. Now its just a pain.

Windows driver foundation pops up straight after a reboot also, Replicated this 3 times over with the same result.

What ive also noticed if it affects cpu clock modulation speeds where it keeps the cpu pegged at full clock speeds without doing anything. Again once dock is removed and not used, behaiviour is normal.



Device is under 11 months old.

Some help / Guidance would be appreciated if possible.

Thanks for reading my rant!

This is happening again all over on windows 1903. I’m going to bring it back up to prompt people to see if they have any fixes.

I have found a work around for you guys. Ill post it up in a bit, just testing it more. Ive decreased temps by 15 degrees idle / working temps. Just testing it before anything else I dont want to mislead anyone.

Alright guys,

So Ive been playing with the said 'workaround' for couple daysd and my temps are back to normal.

My initial finding is that the D3100 clocks the GPU to base speeds, so In my case it would clock it to 1404mhz without having to need to, as I checked use the load is virtually 0 on desktop and even using programs.

Yes the GPU will power all displays end of the day however theres no need for it to stay cranked all the time, generally what ive noticed once clocked up 1405mhz it starts to bring cpu temps up also (as expected).

SO!... Ive been using this tool called nvidia inspector. Link: https://pcgamesbeat.blogspot.com/201...-gpus.html?m=1

Basically I turn it on, turn on muti-display function as described and i just pushed a threshold to 20/20%. And all of a sudden, boom normal temperatures.

Basically what it does is it downclocks the gpu as it normally should (no need for gpu to stay clocked all the way when not doing work. (obviously it should vary).

I have tested it for the past 3 days to make sure it is good and I can still make use of it when it needs to be used at higher power states / function.

This is NOT a fix, this is a workaround if you have a nvidia gpu.

Ive dropped temperatures by 17-30 degrees on both cpu and gpu.

Hope they fix it so we dont have to run this particular thing but for now it allows me to still use this dock.

Any questions id be glad to help, If you guys need I can make a video and upload it explaining more.

No thats fine, mine cut the temps down by downclocking, so what you did there is good. Have you got some sort of factory undervolt? How old is the unit?

Give me some specs

anyone have any luck?

Well its an ATI card and this is a NVIDIA tool, so this probably wont work man.

Of course Im not trying to test this nVidia tool on the ATI graphic card.
Originally one of my HP laptops has nVidia card and that laptop is giving me headaches with WUDFhost, temperature and CPU fan.
After some troubleshooting I’ve noticed that my nVidia card is not in use that much, while Intel 530 is the one used most of the time and that one is being utilized where clock is quite high snd that brings the temperature up while connected to the Dell D3100 docking station.
I’ll have to test without dock and see how it reacts.

Beside that, other laptop I have HP ZBook 15G with ATi card dont have any symptoms when connected to the same docking station.

In context this is what my GTX is doing also, plug it in and it clocks VERY high. Unplug it and it works as per normal. Pass it through to the GPU and use that cool,. or maybe we can find a way of downclocking the IGPU which i wouldnt do.

The Hd530 is on the CPU DIE which is why its running hot (cause its clocking )

Ive spent roughly 80 hours on this issue, Ive also contacted dell about it, dell logged a case and never got back to me its been over a week at this point thats why i try trouble shoot it myself.

I mean we found what the issue is, so least we can go on that man. There is others at my work having the same issue. its a great device, unfortunate thou about drivers.

Hello,

We have created a report to Microsoft, this is with the graphics team at this stage.

There's an important optional OS update for multi-GPU on Windows 10 v1903 currently planned for Tuesday. It would be useful to see if that changes that aspect as well.

Kind regards,
Alban

I will test it and report back, is it an optional update? Such as you mentioned in the other thread?

Also any idea what the update is called so I can check that it is actually installed?

Good morning,

Windows 10 runs on a "patch Tuesday" cycle. Every month, you have 3 updates, generally.
Updates are cumulative.
2nd Tuesday is for security patches and is critical: it is pushed to all machines.
3rd and 4th Tuesday are quality (bug fixing) patches and are optional: only users going to Check for updates will be offered the update.

Because updates are cumulative, users who don't manually check will get the fix as part of the next critical update after second Tuesday of September.

You may see the coding, or may not, I don't remember. The code is month followed by a letter selecting the week in the month. So this update will be 8D (4th week of 8th month), but can also appear as just cumulative patch. Bug fixes are not individually distributed.

Kind regards,
Alban

Will check back after Tuesday to test results after testing the update!

Cheers mate.

I'm running the build 18965 prerelase 190816-1722, still I have the problem with GPU clock.
Anyone has any more details/news on this?

Hello,

We are still in regular exchanges with Microsoft, with different traces being asked.
This is taking longer than desired, but is still active.

Kind regards,
Alban

Hello Alban,

any news on that topic?

Regards
Daniel