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

Sounds good. 3 more weeks, and I will be joining the LGA revolution, too.

 

more like rebellion / guerilla war, cuz of the BGA overmight

Sent from my Huawei Mate 8 NXT-AL10

 

Don't want to sound unpolite, but this CPU paste job on the pic doesn't look good either. That's not enough paste on your CPU. The IHS has to be fully covered. I'm sure it isn't actually.

Try the cross method the next time. Screw it tight, then unscrew it and have a look. If the IHS is covered entirely (without spill over) clean it off and repeat the process. After that you're good to go!

 

The most important area is the shape of the core and then out a bit as the heat has to go through the metal. The rest won't do much, that paste should ge the job done as it follows the alignment of the rectangular core.

 

As Meaker said, that paste job should do the job. In my opinion covering the entire IHS with paste is not a bad thing, but also not essential, since it is not vapor chamber so most of the heat is concentrated in the middle.

As for for the VGA UV I mentioned earlier, anyone else tried it/did some more testing? What is your experience?
In FireStrike I have lost 1-2 FPS despite that most of the times the core clock is now running at 1843-1860 Mhz instead of 17xx-1860 Mhz.
In Rise of the Tomb Raider benchmark I can also see some minor FPS loss (but not during actual gameplay) despite the higher avg core clock.

Is there a chance that with less voltage the cuda cores can process data slower? Even with CPU UV I can monitor some drop in the amount of GFlops processed.

The drop is minor (1-2 FPS) so I still consider it a win because of the significant temp drop.

On a side note, I am having troubles with the latest Clevo Control Center. Today I did benchmarking all day and when I opened the CPU OC app I have noticed twice that the ratio for all cores have reseted to 3x. Very annoying and I very disapprove that the settings made in the app are stored in the BIOS too. Also, what the hell? When I load the factory settings in the bios, why do 3x ratios get set, why not the default 4x ?Or Clevo was lazy to store factory settings for all used CPU types ? When I load the factory settings I expect the settings which my laptop shipped with ...

 

I have been playing around with it and finally understood how the curve related to 'normal' overclocking. If you put in a core offset overclock it just raises the entire curve up by that amount. Which obviously increases the clock speed based on the available voltage. So what I have done is put in an offset of +130mhz and then flat lined the curve at 1860mhz, basically overclocking and then capping which results in an overall undervolt and thus lower temperature, fan speed, and noise level with no performance loss.
I am getting just over 17k on firestrike now after running the demo at the start. Whereas before I would get that on a cold run but dip down to 16.7k ish with the demo at the start due to the higher temps.
All in all very happy with the results!

 

I am currently on BIOS 1.05.01, has anyone tried the latest BIOS, 1.05.02 yet? According to the release notes there are not much changes.

1. System BIOS 1.05.02 for P75xDM3(-G) and P77xDM3(-G) only.
2. Solve “sometimes can’t identify SD card after S3 resume” problem.
3. Refer to readme.txt to update BIOS.

 

Any news of how good you can OC a 1070? I could get a P775DM3 pretty cheap with the WQHD Screen, but im not sure if the 1070 is good enough to handle all the newer Games well, so i want to know what the OC route brings, or I just wait for the 1080 to buy.

Gesendet von meinem Nexus 6P mit Tapatalk

 

I've reduced my vga undervolt to .943v after noticing the odd graphical glitch. I'll try that for the next while and report back. Temps are great though.

I know most of you haven't received yours yet and may think as I did...that this system will be ok and not too hot or loud.
People aren't exaggerating. Without at least a cpu undervolt you may be very frustrated.
I'm only a moderately skilled user and I found the undervolt easy to do. I've also never opened the machine.

 

What kind of graphical glitch?

 

Looks like the 120hz gsync screen has been postponed again, and are still awaiting certification from Nvidia, my reseller through which I ordered the P775DM3 with 120hz gsync screen warned me a few days ago, and I quote; "they are delayed indefinitely, Clevo hope it will be sorted out sometime in November, but they are making no promises . Please, note this issue is in Clevo (and we only received the news today) and nobody in the world is receiving 120 Hz gsync (you can confirm on the forum), so if you get information from another store differently I am sure it is not true. They are having troubles getting Nvidia certification. "

I ended up switching to a P870DM2 with single 1080 and 1440P 120hz screen, I just figured if I have to dish out more for the screen might as well go all the way, not that the P775DM3 is lacking or anything, I just prefer the looks of the P870DM2 .

 

Care to share where you're ordering yours? Most European retailers I found so far have only the FHD screen in store.

 

Well only one glitch actually, a quick fps drop and the main character in the game glitched graphicaly for a second or two. That was it. Like the textures failed or something

 

Bummer, that's terrible news. I've been waiting since the GTX 10 series announcement for the screens and it's now the only thing holding back my purchase of a P775dm2. I feel cursed!

My reseller (Metabox Australia) said they had confirmed shipments last week of October. But it was last week I checked so this is new news.

Is it just the g-sync holding stuff back? I'm wondering if g-sync really matters as much on a 120hz panel. Sure it's good but even with a low fps (say 40ish) doesn't the screen refresh fast enough to help hide tears or stutters anyway?

 

I have traveled a little and I brought the laptop with me to game, however when I have fired up the laptop what do I see? GPU max temp in Rise of the Tomb Raider is back to 88C. Today it even reached 91C when the game crashed...and this is with UV.
I can't imagine what is going on...
CPU temp is around the same however OCCT is crashing with 160mV UV...Lynx (which is more aggressive) passed several times...

 

I'm contemplating getting a single-GPU P870, too. The bulkier build and preparation for SLI means better cooling, at least in theory.

 

Are those temps with overclock fan profile?

 

Yes.

 

Could someone please do 2-3 Fire Strike stress test runs and post the maximum GPU temp after the last run? After the 2nd run I am getting 85C+ temps on the GPU with 0.950 V.

Was also thinking about a delid of the 6700k...does it worth it? After 30 min OCCT stress test (it did not give error in infinite mode) my max CPU temp was 80C with -140mV. (25C room temp)
Now I checked the temps of iunlock after a delid and maybe its just me, but I was not really that impressed.
I mean after same amount of the same test he got 68C max for CPU...if we consider the difference in the amount t of undervolt and room temp I'd say that's about 5C lower than mine.
Unfortunately my CPU can not go as low as -200mV.
But my main concern is not the CPU, since that is not overheating, the problem is the GPU...what am I doing wrong? How on Earth are some people able to achieve lower 70C for max GPU temp? I mean for the first FireStrike run my GPU is under 80C too, but after 2-3 runs its over...if the GPU is stressed alone, it will be in the 60s, however if the CPU is also used (for example while gaming) then the sh*t hits the fan and the VGA temp crawls above 80C...

Should I delid and repasts again OR I should send back the laptop and ask for a better quality heatsink? (Or according to the pictures I have posted there is nothing wrong regarding the quality of the heatsink?) http://forum.notebookreview.com/thr...wingman-2-0-batman-3-0-lounge.794538/page-148

 

What paste did you use again?

 

I get 70C max on GPU regardless of time stress, and over 90C on CPU regardless of settings. Ok, maybe CPU doesn't pass 85C if I limit it at stock 40X on cores, 40X on cache, voltage is ) adaptive, so stock.

With the slightest of OC (even 42X on one core), it easily goes above 90C.

GPU can't heat more, regardless of the GPU bench I do.

Using Clevo driver, all other drivers result in Cinebench corrupted test (try cinebench, cars will show red dots, it will say the GPU is not configured properly).

These temps are constant regardless of GPU settings, Gsync and any other setting, ran many times, same result.

Fan on OC

 

I just have the standard nvidia driver and have no problems at all, I can run cinebench gpu test and get no artifacting and the test completes with no issues.
Your gpu temps are really impressive though! You have the gtx 1080 right? Not 1070? I tried the clevo driver in case it was something to do with that but my temps were the same as with the stock nvidia driver.
For the record my setup is currently -125mv on cpu and 950mv max gpu and oc fan profile and my temps after a couple of hours playing the division with gpu maxed out constantly and it is a pretty cpu intensive game I got max gpu temp of 81 degrees and max cpu temp of 78 degrees.
I ran 4 firestrike runs back to back and got max 76 degrees gpu and max 73 degrees cpu.
These are with the laptop on a desk, if I use it on my lap I get lower temps by around 4 degrees due to better air flow.

 

Did a run of Fire Strike stress test, ambient 22C, -140mV CPU undervolt, GPU stock, overclock fan profile. Peak CPU temp was 78C, peak GPU temp was 75C.

 

Noctua NT H1, one of the best. However it was sitting in my desk for years since the last use.
I was thinking that it might have lost its thermal conductivity after a few heat ups/cool downs...but then I was thinking, can this happen with the previous Noctua paste and this one too? (have no idea what the precious paste was, service only said that its Noctua)

 

indeed, pretty much on par with IC Diamond

http://forum.notebookreview.com/thr...e-recommendations-update-thermal-pads.796820/

 

I tested Noctua Nt-H1 on my Oc'd 4930mx. A big fiasco!! See also results from @Mr. Fox testing of this junk compared to ICD and Liquid ultra.

 

NT-H1 was not good for me at all.

 

A good paste for first timers. Very Easy to apply, but this junk pump out in short time. Especially if you Oc.

 

For machines on the go a lot use a thicker paste like IC diamond to avoid pumping.

 

Gonna buy this today and try it out:
http://www.coolermaster.com/cooling/thermal-compound/mastergel-maker/

IC Diamond is not available in shops here (or its rare) but hopefully it will give about the same results (this is filled with diamonds too)

 

Regarding the clevo cc cpu overclock feature.

When i undevolt to -150mv and set multipliers to 42x, sometimes the power limit changes from 95w down to 50 something.
It seems to me that having the limit at 95w max wouldnt mean that the undervolt would have a problem. Why does control center lower the power limit? ( note that it doesnt do it right away, usually after a few restarts i check and its happened)

 

That program is really buggy. It would be better to use ThrottleStop. Even XTU would be a better option than the Clevo OC utility, but @unclewebb has given us something truly great in ThrottleStop.

 

Yes, id rather use throttle stop but i thought that i would have to uninstall clevo cc and xtu to use it properly. Id rather not lose the overclock fan profile if I dont have to. Is there anyway to use xtu or throttlestop and still have cc installed?

 

Yes, just don't use more than one of them for overclocking. Reset the Clevo OC utility so those settings do not interfere.

Do you have the older, less buggy version of Clevo Control Center already? If I remember correctly what others are saying, the version for the pre-Pascal models works better than the newest versions.

 

I think it .78 The one thats downloadable off the clevo site for the 775. I'll set it up with throttlestop and see how it goes. Thanks

 

That could be causing some issues. I think you need to use the one for last year's model instead. maybe other owners of that machine can chime in and share what they did.

 

Hmm, throttlestop is doing the same thing. Under a stress test it is tdp'ing at 53w. The same wattage the cc was throttling it at. I dont get it. That makes an undervolt kind of useless and give me slight stuttering in games everytime it throttles. How do i increase the tdp permanently in throttlestop?

 

You can Increase all limits in TPL tab.

 

thats the disadvantage of having raw benchmark numbers: they dont signify the long term performance of a paste

Sent from my Huawei Mate 8 NXT-AL10

 

HaHa. But nothing is more fun than see bench numbers fly and the temperature rise A lot of people expect that all thermal paste works equally good in notebooks as desktops. They forget that hardware temperatures in laptops is normally higher than desktop hardware. + the heatsink is of less quality and also haven't the same high pressure on processor IHS/die. If laptops had better heatsink, would the situation be another game. The worse cooling in notebooks ruin thermal paste a lot faster. But people forget this And save $ 3-5 on thermal paste isn't very vice. This isn't the place you should save money.

 

Everytime I make voltage offset changes in cc it changes my pl2 limit to 43w or something. This then updates the bios to the same, so eveytime I have to go into the bios and manually change it back. I guess that is what's happened here: cc has updated your bios to the lower power limit. But you can change it back and then I guess you will see it updated in throttlestop too.

 

Ya, I got frustrated with throttlestop and tried xtu. ( not through clevo cc). It made me restart system to allow power level adjustments to stick. Clevo control panel seems to let you change the power limit but it doesnt do what xtu needs ( a specific restart) to allow it to stick. Things look alright now, my bench tests are good and hopefull gaming will be smooth and cool!

 

Just disassembled my laptop and I have to say, I am not proud of my previous paste job...
http://i.imgur.com/RdUJQg1h.jpg

 

Heh. You might've as well used no paste at all .

 

Are those pads stock?

 

Yes, they are stock.

My new paste job:
http://i.imgur.com/4RBf0QDh.jpg

First time for me to use a paste like this, I hope its acceptable

 

mine has the same pads I don't like them but the amount they're squished down reassured me they're at least the right thickness

 

A little reminder, my results with stock paste:
Ambient temp: 22C

CPU: -160mV UV, no VGA UV
Max CPU temp (Running 20 rounds of 4096MB Lynx CPU stress test) 88C
Max GPU Temp: 1 Fire Strike Stress Test: 85C
1 hour of Rise of the Tomb Raider 88C

Here are my result after my new repaste with nano diamond filled Cooler Master paste:

Ambient temp 27C

CPU: -170mV UV, VGA: -112mV
Max CPU temp (Running 20 rounds of 4096MB Lynx CPU stress test) 80C
Max GPU Temp: 1 Fire Strike Stress Test: 77C
1 hour of Rise of the Tomb Raider 75C

Fans were on "Overclocking" profile during testing.

Before repaste +VGA UV my results in Fire Strike Stress Test were 95-97.3%, most of the times the VGA did not pass the testing because it was throttling due to the high heat. After UV and repaste my results so far were 98.3% and 98.9% (you need 97% to pass the test).

Now I just hope that this new paste will not loose thermal conductivity in a few days.

 

Ideal thickness is a light impression on the surface. Might be worth investingating.

 

I would get some new thermal pads... Look for Phobya XT ones, they're pretty cheap and good! Those stock pads look ****!