*** Official Clevo X170SM-G/Sager NP9670M Owner's Lounge ***

Single core should be able to turbo, that is odd, have you reached out to scan for their update?

 

Anyone who has this can confirm nvidias whisper mode is working? X170km-g seems it's disabled.

 

Yes, I have. Hopefully they can suggest something. They don't have any driver downloads section for these machines so I'm assuming they just use the ones on the Clevo web site and there is a new BIOS there. Just wondered if there was a known issue with 1.07.03 clamping down on the CPU compared to 1.07.02 from the first chassis I had.

 

We hand out bios files to our users on request so that we make sure they are flashing the right one.

 

Just for fun, they are all almost the same size......

 

At least in area

 

I just started using rockit's copper IHS and liquid metal both between die and heatsink, I did not seal it back considering I needed to check and test, however, the cpu still stubbornly thermal throttles ( under combined full load) and it actually gets worse results than the normal setup with thermal paste ( 10600k 2080s). Is it possible it's not making proper contact for some reason?

 

Absolutely, liquid metal is very unforgiving of any angle on the contact.

 

I will have to buy another bit of LM anyway, wasn't sufficient for both cpu and gpu, it is kinda spread thin, so maybe it is that as well. I checked the contact pressure and contact zones many times, and it does seem to make contact in the right place, maybe I'll try again with more LM and a cleaner surface

 

I have been using my machine with 240 Hz Auo screen panel for quite long hours on daily basis as I have been doing my office work on it as well. I have noticed that every time it gives me severe eye fatigue when I have been using it for 3-4 hours or more at 240Hz refresh rate. I then decided to test it at 60 Hz and boom! My eye fatigue problems are gone just like that! Has anyone else noticed it with these screens?

 

No, but you might be sensitive to flickering or similar quick picture changes. You have epilepsy of some kind?

 

Nope, nothing. Perfectly healthy. I think that these screens are using PWM for screen brightness control as compared to DC voltage based dimming which is a known reason for eye fatigue. My suspicion also gets reinforced about it because I have noticed that at full brightness, I do not feel fatigue however in that case I have to make sure there is enough natural ambient light in the room otherwise it is tough on the eyes to look at such a bright screen for long durations.

 

All of the "Copper IHS's" that I have checked for 10th gen chips so far DO NOT match the thickness of the stock IHS which explains why you are having worse performance. Move back to the stock IHS and you will thank me later. Ditch the copper IHS in this setup because the Z-height is a VERY sensitive matter in a laptop chassis, unlike a desktop where there is a separate cooler for the CPU alone. I have tried Rockit Cool copper IHS for 10th gen myself and let me be the first to tell you that it is THICKER than the stock IHS by a significant margin. If you have a precision micrometer screw gauge, you can verify what I am saying. This will make it difficult for the heat sink to sit flush on it as well as the vrm thermal pads making good contacts all around the CPU. When over heated, VRM output voltage curve bcomes a lot more noisier and that too adds to the overall thermal output of the component they are driving. Keep the copper IHS for some Desktop build like I have done.

 

Screen Flickering / PWM (Pulse-Width Modulation)
ℹTo dim the screen, some notebooks will simply cycle the backlight on and off in rapid succession - a method called Pulse Width Modulation (PWM) . This cycling frequency should ideally be undetectable to the human eye. If said frequency is too low, users with sensitive eyes may experience strain or headaches or even notice the flickering altogether.
Screen flickering / PWM not detected
In comparison: 51 % of all tested devices do not use PWM to dim the display. If PWM was detected, an average of 9724 (minimum: 5 - maximum: 151500) Hz was measured.

From the review of this device.

 

It seems that our brain which is directly connected to the retina via retinal nerve, does not like the very rapid pulse width modulated light signal because there is always a sharp impulse involved in artificial pwm controlled power signals. On the other hand, all natural signals are always sinusoidal in nature. This is one big reason why pwm almost always causes problems. For now, I am living with it by keeping the refresh rate at 60 Hz all the time except for when I want to game which is not very long. Once done with gaming, I jump back to 60 Hz. My eye fatigue is much lower now than it was before (but it is still there). However, I have a separate monitor that uses dc power to lower the light intensity and it has never caused me any eye trouble (I have been using it as my main work monitor). By the way, I noticed the exact same problem with Dell's Area 51m 1st gen screen which was 144Hz. Apparently it is cheaper to implement PWM controlled brightness at mass scale than it is to use dc voltage dimming. What a shame.

 

This makes me really want laptops to have separate heatsinks for the CPU and GPU again like we had 5 years ago.

 

And here I thought it was my mistake, I could see it wobble left to right when only laid upon the cpu die....well, thanks man, I now surely have to buy more LM. Which one would you regard as the best option? I used conductonaut so far.
Also any good nail polish or similar product you trust ?

 

I use StarBrite Liquid Electrical Tape (DON'T INHALE! The stuff is TOXIC!) to seal the contacts next to the CPU die however if you have a nail polish, any one should do the job just fine but you will have to ensure that you coat those contacts thoroughly with at least 2-3 coats of the material and letting it completely dry between the coating intervals. As far as LM goes, conductonaut does the trick. I bought 1 gram syringe almost 3-4 years ago and there is still some of it left today lol.

 

It's usually not a lack of liquid metal, you really don't need a lot..

 

Speaking of delidding...

10850K is on it's way back. Stock IHS instead of copper, so looking forward to testing it. The i7-10700K it's replacing is going to be delidded with a copper IHS since that'll be going into a desktop. Hoping the temps will be a lot better.

 

Wait you are going from 10850 to 10700?

 

Other way 'round

 

Nope.

Laptop is a 10700K. That's getting the 10850K.

Then the "spare" 10700K is being delidded, going into a desktop motherboard and going in my PC, then the Ryzen 7 2700X from my PC is going in my wife's PC to replace her aging 5820K.

 

Ah nice, yeah up to around 15c with liquid metal and good IHS.

 

So i've heard And that'll be going under a 280mm AIO, so temps should be good.

 

On desktop yeah that won't pose an issue at 280.

 

Well, the 10850K is in. Ooof it's a warm boi, but also running slightly cooler than the non-delidded 10700K.

Also undervolts better. So I'm quite pleased with that.

Going to take a trip to the hardware store on monday to get some better washers as I think mine aren't fitting properly.

EDIT: Scratch that. Amazon had a box full of washers and I had a gift card, so that was an easy decision.

 

Always good to have Amazon on hand for that sort of thing

 

So i've got an odd problem i'm trying to diagnose (and have been for the past 24 hours).

System loads windows fine, everything runs fine. I can watch youtube with no problems, twitch, run unity, code, etc.

If I try and play a game, it works fine... for 5-10 mins. Then the USB mouse starts to stutter and not respond (eventually just not responding to me at all, but the touch pad works), THEN audio starts to play super slow and crackly. Temps are good (80-ish on the CPU, 65 on the GPU) so I don't think it's that.

BUT... if I reboot into Linux and play games on there? No problems at all. I've already done a windows repair install (no change) and a completely clean windows install (no change).

Any thoughts?

EDIT: Think I fixed it. Even though I did a clean install with clean drivers, I grabbed the latest ones off the clevo site and reinstalled all the drivers AGAIN and it seems to be fixed now.

 

I haven't been on thread much but wanted to know if anybody has tried the new heatsink for the rtx 30 series in the x170sm? It looks exactly the same just with wider holes for gpu. I still have my x170sm. Friend of.mine got a hold of a 3080 for me but told me I needed to purchase the heatsink. Fortunately no one purchased my laptop that I had for sale and someone blessed me with a stock that sky rocketed and I was able to solve my issues. The heatsink I have found is extremely expensive priced at 400 bucks. If anybody know who sells the spare part cheaper and it fits let me know. @Donald@zTecpc @Mr. Fox @Papusan @S.K thanks all I appreciate it and sorry if this has been answered on the thread already there's literally over 250 pages I've missed. Also if anybody could give me a fair price to sale my 2080s for I'd appreciate it

 

XMG stated theres no uograde kits available

 

Due mostly to the GPU shortage, Clevo has not released an upgrade kit, and says not to expect one for months.

 

Ok, the issue is definitely still not fixed. I'm at my wits end.with it.

I managed to resolve the mouse issue. Not sure how/why, but I swapped for a different one and that's now working. The sound issue is 95% fixed. If I turn turbo off (set max processor state to 99% in windows) then sometimes the audio issue occurs, sometimes it doesn't. If it DOES occur, it's after 20 minutes instead of 6-10. Temps were 75-77C on the CPU when it started (GPU at 65). Exiting any programs and leaving it for a while restores the audio to normal.

I'm going to try reinstalling linux later and see if the issue is there, but it wasn't before I nuked the boot drive. I'm honestly stumped. I may try reinstalling the audio driver and see what happens just because i'm out of ideas.

 

If you are using DCH video drivers that might be causing problems. DCH trash drivers are known for glitches and memory leaks.

Use NVCleanstall and DDU. Download non-DCH drivers using NVCleanInstall and tweak as desired. If you don't use GeFarts Experience, don't select it. Save the modded driver package to a new location. Nuke the DCH filth with DDU and then install the non-DCH drivers.

 

Non-DCH drivers won't install. Keeps saying I don't have a compatible device.

 

If you have not fully removed all traces of DCH drivers (including the UWP NVIDIA Control Panel) with DDU that error message will always occur. You cannot install non-DCH over DCH and a normal uninstall procedure (including NVIDIA's own "clean install" feature) is not adequate. Run DDU, use the advanced options and remove everything. If you are not using it in Safe Mode, then choose the option to shut down (install new video card).

Disconnect from the internet completely before starting the process. Sometimes Windows Update will bugger up the process by "helping" you install the DCH trash. That can also produce the same error message. Once you have non-DCH GeForce drivers installed, Windows Update should no longer attempt to update video drivers (unless something has changed).

 

That's exactly what I did. Downloaded the driver, downloaded DDU, disconnected from the internet, rebooted into safe mode, ran DDU, told it to remove ALL traces of nvidia, ran the drivers I got from NVCleaninstall, told it to run with what IT recommends, got told I don't have a compatible device installed.

 

I am not sure what is wrong. Maybe download a new package with NVCleanstall. This time though you manually select a driver version (not what NVCleanstall recommends) and confirm you are downloading a non-DCH notebook driver for W10. Maybe a desktop driver was accidentally selected before, in which case the HWID for your system will not be present in the INF files. You can also use NVCleanstall to manually add a HWID to the INF for older GPUs that are no longer supported by NVIDIA. Just to be sure Windows didn't sneak something in again when you were not looking, run DDU again before installing the new driver.

I did this multiple times with the X170 that I was tweaking for zTechpc, so I know it works.

Beware of the latest GeFarts drivers though. I would choose one a release or two older. About a week ago I started having trouble with one of my desktops (the one with the 2080 Ti). It was totally shutting off when DX12 was utilized. It would abruptly power off as if I unplugged it from the wall whenever a game with DX12 was launched. I could run 3DMark benches like Fire Strike and 3DMark 11 just fine. I dropped back to an earlier driver and everything was back to normal.

 

Sorry, should have specified. I manually selected the latest, non DCH notebook driver, but when it asked what components I wanted, THAT is when I clicked "recommended".

And it told me I didn't have a supported device.

I think I've fixed the problem though. I grabbed the Geforce studio drivers instead of the game drivers, installed those, problem remained (best way to describe it... you know when you tune an old radio so you get static and you have to keep fiddling until the voices fade in? EXACTLY like that), looked around online and found a similar problem for a racing game that suggested changing all audio formats to 16-bit, 48000Hz (DVD Quality). Did that, then found a post about my specific audio interface that suggested increasing the ASIO buffer from 512 to 1024 and turning up the buffer size so the CPU is used less.

Did that too, just recorded for 30 minutes, no audio issues.

 

out of curiosity, what are your results with LM applied ? (2080s) talking mostly when going full throttle on both cpu and gpu, so around 360 370watts of consumption

 

The heatsink i can purchase online right now they have 8 available. The 3080 a friend of mine acquired he has good friends in clevo and in china in general he said he will have it in hand 5/19/21

For heatsink look up 6-31-X17KN-100 on aliexpress i feel its over priced but it is what it is

My friend also let me know that a water cooled heatsink will be available by the end of the month for the X170SM which will support both 20 and 30 series! @Donald@zTecpc this is the surprise i was holding on to from back then. Took longer because its in final phase of testing. He also said it has a small improvement on air cooling the cpu but not to hold my breathe for dramatic improvement.

 

That sounds awesome!

$400 for a heatsink is way overpriced. It should be more like $150.

I'm looking forward to the release of that water cooling heatsink though. Do these custom water cooling heatsinks for the Area 51M, P775, and P870 still work at least as effectively as the stock counterparts when they are not filled with water?

 

Hell, $200 would be reasonable as well. I may consider picking one up in prep for upgrading the GPU since i'm just running a 2070 super, so a 3070 or 3080 would potentially be an upgrade.

 

But, we live in a day when literally nothing is reasonable. Unreasonable people, unreasonable prices, unreasonably low standards are becoming the new normal.

 

Eventually supply and demand will normalise.

 

Sadly very true

 

The heatsink is much cheaper, the chinese thief/reseller just takes a huge cut.

 

Yeah the manufacturing cost is the same as usual on the heatsinks.

 

Yes, prices are practically exploding. Everyone thinks we wait a bit now and it will be like in the past that everything becomes cheaper again! I have wondered for years when the exorbitant financial market returns and bubbles have an impact in the real economy. Now the time seems to have come!
I have times my model calculation made from the statistics of the Bank for International Settlements(BIS)...It is air about 99.987% in the system by the interest main....

https://stats.bis.org/statx/srs/tse....5J.A.TO1.TO1.A.A.3.C?t=d5.1&c=&p=20201&i=1.4