*** Official Clevo P775DM2/3(-G)/P75xDM2(-G) (Sager NP9152/NP9172) WINGMAN 2.0/BATMAN 3.0 Lounge ***

Black ops at 3440x1440 really taxes the gpu. The frames are 100 but the Temps are between 82and 92 even at maximum fans. ...and I undervolted the cpu to -150mv.

I think I need a repast and I'm scared to do it. Maybe that gelled stuff. Geez, I thought the ic diamond they put on at eurocom would be ok

 

Maybe the thermal paste application isn't done properly or you got a **** heatsink. Or even worse... Both sucks!!

 

After some research and conversations with multiple people I have decided to go with the new build and sell my current rig. However I am going with 1070sli over the 1080sli. Simply not dealing with the heat everyone is seeing with these 1080s. I will likely have the laptop for 2 years so the 1070s will perform fine at 4k until next upgrade. Looking to place my order within the month.

 

I wish i'd gone with 1070 sli in my 775DM3. ​

 

What thermal paste would you personally suggest?

 

Got the laptop back yesterday with a new screen and a new paste job.

What are the temps you guys get with a 6700K and GTX 1080? My temps seem to be the same as with the old paste job, meaning that een after -160 mV CPU UV (this seems the most stable for me) I get 80C max for CPU and 87C max for GPU after playing 1 hour of Rise of the Tomb Raider.

I have Noctua NT-H1 paste home, maybe I should try re-pasting it myself ?

 

http://forum.notebookreview.com/thr...dup-performance-price-recommendations.796820/

 

I use Liquid Ultra. The only choice... For me

 

Anyone been hands on with the 1440p screen yet?

Also, it's confirmed that the 1080p, 120Hz Clevo has is an IPS?

 

Yes it's an IPS. Unfortunately I won't be getting the 3K TN screen as I don't play FPS games so much.

Oh, did I say that I pulled the trigger?

Hadriel

 

I need the AUO model number. It'd be a great addition to my current laptop.

Anyone have the info?

 

@Papusan. ( or whoever)

If undervolting a cpu doesn't affect clock speed or performance..then why is it a bad idea to undervolt the gpu? (That thing is smokin!)

 

It could make the CPU unstable if you are not careful.

 

I undervolted the CPU in Clevo software and resintalled windows. The settings made in the Clevo software in the previous Windows has been stored in my BIOS and on the new windows the "CPU OC on" is on by default, however I can not access that part of the software. I double click and loading screen appears and after that nothing. GPU OC and main Clevo software opens without any problem, only the CPU OC part which seems to now work. Maybe a bug caused by the settings read from the BIOS? Anyone experienced something like this? What did you do?
I did a "Load factory defaults" in the bios, but the core ration is all screwed up now, it is 31-30-28-28 now and CPU OC part of the Clevo software still does not open. Any idea ?

 

I've used liquid metal and have dropped my temps over ~20C+. (Grizzly Conductonaut)

IMO, with how hot these beasts run, traditional paste is not even an option for me.

IMO, I do not like ICD at all and think it was a cheap escape goat to cover / fill up the intolerances of warped heat sinks and/or due to it being one of the thicker pastes, which is another issue in itself. (Long story short, more fillers to bind the diamond dust to hold it together, thus at the expense of lowering the w.m/k potential.)

I also do not favor how ICD scratches up the die of the chips. Frankly, it's just annoying to know that it's there. (Ex. It's like buying a brand new car, taking a razer and slicing the lights out of your leather seats and accepting the fact that a seat cover will hide the blemishes...it'll hardly ever be seen nor would it impact the comfort, but the tears are still there.)

Now despite my opinion, what holds to be true is that ICD is rubbish as we have seen over and over again how it can not do the job properly to keep the beasts cool. It just does not have the capacity to do the job well and that also holds to be true for most traditional pastes. The only traditional paste that I'll use are Gelid Extreme and/or Grizzly Kyronaut if I had to.

This is why with these machines, using liquid metal to provide the maximum thermal conductivity potential is IMO the only sensible option. We're already dealing with many facts such as having desktop components in a compact environment with tolerance issues, lack of QC, pastes that's not good enough for what's at hand and uneven thermal pads cut and applied by incompetent people.

For anyone who's new to repasting, I'd recommend focusing on replacing or realigning your thermal pads first making sure you have a proper fit, then maybe use either Grizzly Kyronaut or Gelid Extreme...(wear latex gloves as you don't want to get the oils from your hands on the pads.) - After this first oil change you'll get a good feel of what to do without dealing with the risks of getting liquid metal on your mobo during application.

Then, perhaps go with CLLU liquid metal as it is a bit thicker and easier to spread than Grizzly Conductonaut.

::iunlock::

 

You can with a mod vbios. But more risky. A reset of graphics settings isn't doable like processors in Bios if you run into troubles. A few here run graphics with undervoltage, but not many.

Try a Nvram reset(pull out cmos battery a short while).

 

That, is a very good reason not to undervolt a gpu. I was wondering. I guess I can't anyway without a prema vbios.
Well..i'll have to deal with this heat another way

 

The AUO 120 Hz IPS screen is supposed to be AUO B173HAN01.1 (based on Eurocom's website).

Hadriel

 

Try manually opening it...

C:\Program Files (x86)\Hotkey\CPU_DRAM_OC.exe

 

Reinstall CC . Had this happen. There are many workarounds for it. the idea is that if you don't have XTU installed, or it's not installed correctly, CC CPU OC won't work right.

Start in safe boot, uninstall CC with geek uninstaller, install CC again, it should be fixed. Other option is to reinstall windows. Don't run it without XTU, CC CPU OC won't work without it's own XTU

those factory defaults don't go under 4.0GHz on CPU.



1080 never gets above 70C. It still scores well enough in benchmarks and is ~1-2% above most 1080 scores. No idea what is going on. No mods, no OC no voltage, no nothing. Too afraid to play with it ;_;

 

Seems you have a good paste job, or that your chip is a golden chip? I'm pretty sure undervolting would be fine, but if it's ok then don't touch it, until you need an OC.

Once I get a final confirmation from my side, I will let you guys know what I got.

Hadriel

 

Thanks guys!
Actually the fix described here worked like a charm:
http://forum.notebookreview.com/threads/clevo-control-center-oc-cpu-ram-app-not-working-read.796478/

 

Can you please run a FireStrike test and post the score+max GPU temp after the test?

 

Trying to repaste the CPU and GPU and removed all 9 screws but the heatsink just wont come off. Am I missing something or I just need to apply more force?

 

Have you seen my threads with screen shots running OCCT and other benches? Here's one and here's the other.

On stock clocks after running Fire Strike just now,

GPU Temp is: 64C

 

Might be sticky paste. Try some more force, but not too much. Don't want to pull a fender bender on your pipes :-D .

 

Well repasting in progress, I have managed to get the heatpipes off. Before repaste mx GPU temp was 81C after 1 3DMark FireStrike run. (CPU -160 mV UV)

Progress so far:
Old paste #1:


Old paste #2:


Heat spreader quality after cleaning old paste with alcohol:


Not the best quality I saw regarding copper heat spreaders so far ...

 

What paste will you be using?

As for the copper heat sink plate, mine didn't look as tarnished? ie...the fade marks. Hmmm...

Your thermal pads look to be making great contact. Very squished haha...mine were / are the same.

 

An around 5 years old Noctua NT-H1 lol

 

Good luck with that ****

 

Eeeks...mmm...keep us updated

Order some CLLU mate. It's worth every penny. Just be thankful that repasting the beast is a breeze.

 

Finished repasting. Some more pics:

Cleaned CPU/GPU:

New paste (5-7 year old Noctua NT-H1):


An interesting label on a part (not a big deal, just funny for me):


Ran 2 Fire Strike test.
With the old paste for the first run max GPU was 81C.

New paste:
1st run max GPU temp 74C
2nd run max GPU temp: 77C

Doing a killer burn in test now with unlocked frames in Metro 2033 Redux. With old paste max GPU temp here was 88C after ~20 min.

 

Run OCCT for 20mins when able. Curious to see the temps.



::iunlock::

 

Are you running at 1080@ resolution for those temps?

It occurred to me that I'm too harsh on this system for temperatures considering I'm running either 4k or 3440x1440 and hitting the 80's with Max fans after 15 min of aaa gaming probably isn't that bad for a high end desktop rig smushed in a sardine can.

I think I'll try some stress tests at 1080p to see how significantly lower the temperatures would be on the gpu

What I'm really thinking now is how to adapt some refrigeration or the like into a laptop cooler

 

it does not matter if its 1080p or 4k, what matters is if the game is using the GPU 98-99% for a long time, the temps reach the 80s.

Just finished my test and after the repaste it seems my temps are about 4-5C lower. Will still go above 80C if a game is using 99% of the GPU for a longer period. The temp will be in the 70s for the fist few minutes but as heat build up inside that small space, it will go over 80C.

A good example is Metro 2033 Redux vs Rise of the Tomb Raider. Both have really nice graphics, however if G-Sync and 75 FPS limit is active, Metro is using about 60% of the GPU and the GPU temp will not go over 70C while playing this game, however in Rise of the Tomb Raider after 15 min of playing I go over 80C because RoTTR is using 95-99% of the GPU. Yes, both are running in 1080p. (If I disable the FPS limit, Metro will use 99% of the GPU too making the temp go over 80C).

Georgel: Thanks m8! What driver are you using? Your temp is excellent, 7-8C lower then mine, however your GPU score is ~1000 less then mine.

 

My FireStrike temps after more than 1 hour of 99% GPU stressing (with Metro and Tomb Raider):

 

Driver that is inside Clevo's site. With all other nvidia drivers I get crashes and such.

I have a problem. with almost all mouses I've tested, my mouse jumps around hectically for low and fine movements, while touchpad and digital drawing pad are doing fine movements just fine.. Wonder if it's a driver, a video driver, a hardware, or a mouse problem ;_;

About scores, it's bothersome that I can't get over that score and similarly bothersome that it won't get over 70C for me regardless of what I do...

 

CPU is touching 90C in the same test, locked at 4.0GHz, after running multiple tests, running the CPU at a lower frequency yields better results. Maybe it's a cooling thing...

 

Maybe I am just paranoid, but is there a chance that the driver form the Clevo website sets a 70C temp limit resulting in downclocking when the GPU reaches that level? This could explain why it does not exceed 70C no matter what and also would explain the lower graphics score.
As for CPU temp, dont forget that I have a -160mV undervolt.

 

Terribly optimised engines such as CryEngine that plaster the screen with ****ty, resource-hogging post-processing effects will always strain the **** out of the GPU. So I wouldn't worry too much as long as you can play DOOM at ~140 fps without turning your keyboard into a barbecue .

 

Just did a little VGA undervolting of -137mV. (CPU got -160 mV) (For the VGA 0.925 V is the highest now instead of 1.062 V)

FireStrike run:


My con/pro so far after 1 hour use:

Pro
- temp has dropped ~5C
- FireStrike Stability test gives a higher score compared to the factory voltage (88.3% vs 87.5%)
- higher boost clocks (1860 Mhz can be hold for a much longer time) during gaming resulting in around the same or slightly higher avg fps

Con
- graphics score in FireStrike dropped about 140 points

And finally, my GPU temp after 1 hour of Rise of the Tomb Raider (as a reminder the temp before repaste and VGA UV was 88C)

 

How did you undervolt the gpu?

 

Download the latest MSI Afterburner (Beta14) and install it.
Open the program and press ctrl+f.
In the new window with a curve set 1860 Mhz for 0925V. This way the highest boost clock during usage will be 1860 Mhz with the highest voltage of 0.925 V. You may choose other settings, the reason why I chosed this is because 1860 Mhz is the factory boost clock and as for voltage, I did not dare to go lower.
Apply settings and test.

So far I have not been able to pass 74C and this voltage seems stable. So far it seems FPS is around the same in games, the only difference is the slightly lower graphics score in FireStrike, but I think -14C VGA temp > 130 points extra in FireStrike.

 

How do you test stability with 3Dmark ?

 

There is a stability test option

 

Just given this a quick try with the same voltage you suggested and yep seems to be completely stable and just like you I cannot get above 74 degrees gpu temp. Currently testing on the division with unlocked framerate. As you say this also gives a more stable clock speed because the power limit isn't constantly making it drop.
This is a great feature especially for laptop owners as we can now drop our gpu temps significantly along with cpu by backing off the manufacturers extra safe voltages.

You get my vote for winning the internet today

 

Dude, seriously. When I get home I'm going to try this.
It sounds like you deserve hero status or something because if I can play under 80c without those darn whiny fans blasting I'll #*+\% my §±%~¿¡!

 

You can lock in a specific frequency/voltage with ctrl+L too.

 

Amazing improvement in Temps using the beta afterburner voltage curve.

Easily a 10 degree Celsius improvement in gpu Temps. Still 63 frames per second @ 4k. That was with automatic fan...so quiet and cool. Mid 70's and cpu was abit lower.

I guess I'm set from this point on, enjoy ur new beasts!