*** Official Sager NP9175 / Clevo P775TM Owner's Lounge! ***

I noticed some strange behavior from my 1080gtx.

When I'm playing Witcher3 I set my 1080gtx to a conservative curve undervolt of 1835mhz @ 950v and my cpu to a very conservative 3.7ghz all core (-125 undervolt). RTSS capped fps at 60. This allows me to play W3 with ultra settings (also hairworks & HBAO+) at a constant 60fps while my gpu temps remains under 70°C (!).

Yesterday I was playing for over 2 hours, when suddenly gpu temperatures began to climb from 70°C to 90°C in a minute or two. The game hang and crashed. Temperatures dropped back to normal. Investigating gpu behavior I saw no voltage, use, or power spikes that could explain the sudden rise in temperature. Fans were close to max speed and didn't decrease. Restarting W3 showed normal behavior.

 

Sre the passmark for boards that delimit power out of the box though? Compare with gamers nexus numbers.

 

The contact between the IHS and the heatsink is good. I think it's probably a bad liquid metal application, how even HIDevolution resealed the IHS and i do not have a Delid tool.

I'm not running a override voltage. I'm using -90mv offset. The 1.2V is just what is read under full load.

Yea, future frame rendering is on because if it's off in DX11 I lose about 50 fps. The thing is I am using adaptive with -90mv offset. Any lower and its not stable.

 

Have you tried pushing the undervolt offset further? -90mV at 4.4Ghz seems very modest.

 

I've tried -100mv doesn't get past windows load.

 

So, heatsink contact with IHS is good and the paste spread is alright. Then I suspect the liquid metal application between IHS and die, especially since you mentioned it was resealed. Even a thin layer of glue does slightly add to the overall height of the IHS, which in turn results in a larger gap between the die.

Is your 8700k a binned chip or?

 

Not it was not a binned CPU. Even then most 8700K should do 4.7GHz at reasonable temps. According to HIDevolution, during their test at 4.8GHz max temps they got were 90C with -85mv. That's not what I'm seeing though, so I don't know.

 

I'd reach out and explain your issue, the HIDevolution team is great and I'm sure they can help you out.

 

Hi Everyone, I hope you all are well.

I am just after some advice. I just recently purchased a P775TMG with the following specs

- 9900KF (-120 under volt)
- RTX 2080
- 32GB 2666MHz in dual channel
- 970 1TB Samsung Evo Plus M.2 PCie
- 4k 60hz G-Sync 100% Adobe RGB
- Prema BIOS

My first unit had a faulty MB(USB C and Thunderbolt was not working at all) and so they completly rebuilt me a whole new system so I still use the current one until my new one came. Then on my new one there was a dead pixel which I understand these things happen so I got that replaced.

Now after a week and half with my new unit I am really happy with it. I love this thing. I can lug it around to work and also use it for gaming.

But I just realised and I swear it wasn't there when I first got it but I can hear tiny bit of coil whine from the GPU just before the fans spin up when I am pushing above 150+ FPS.

This machine is preforming about 13% better on the GPU then my previous unit as it was only the thunderbolt and USB C that was not working. I really don't want to RMA it again but I was wondering if anyone elses RTX 2080 or RTX card has a bit of coil whine?

Thank you for your advice in advance.

 

Coil whine is normal. I wouldn't worry to much about it.

 

There are a lot of factors that can cause coil whine like humidity, temperatures, load etc. Unless it's really obnoxious I would not worry about it.

 

Yeah I think your right, I am just being overly picky as my first unit I had some bad luck and this unit had a dead pixel.

 

It isnt that obnoxious at all you can't really hear it when the fan is spinning over 50% while gaming and it is also is only when I am pushing above 150FPS

But I am happy to take a small amount of coil whine to get 13% increase in preformance on the GPU and I have not overclocked it haha

I am new to this I have always had MacBooks but I read somewhere you can undervolt the GPU and it sometimes can make the coil whine qieter or remove it?

 

my P775Tm with 9900kf crash, hard shutdown if i undervolt more than 50... my power limit PL0 - 65w PL1 - 95w. in HWinfo Cpu Vcc - 1,41 volt maximum (it is with 50 offset). i benchmark linpack, in this settings more stable, if i downvolt -100, crash after 5 min - BSOD ... Failury Power status... and HWinfo fixed hardwsre errors- CL0 Errors.
please, say what your power limit in bios, voltage with max load, temps with max load and revision your CPU. i have P0 revision.
maybe bad version cpu i buy new

 

I am sorry to hear but some processors can be underclocked more then others. It is just if your lucky I guess. My undervolt is adaptive is yours?

My power limits are:
Power Limit 1: 120W
Power Limit Time: 28 seconds
Power Limit 2: 165W

 

It is my default setting was. But cpu is throttle fast - 5 sec temp 95C and thermal throttling.
Adaptive undervolt - 50mv. If 100mv - system crashing, BSOD, Hardly turn off.
Can you make test in Cinebench R20? And monitoring Temp while going benchmark?

 

Hi guys, does anyone know what thermal pad size for the 1070(TM1 version)? Thanks.

 

Ops

I did Cinebench R15 and I got 1947 as the score after I kept running it multiple times.

The temperature did get fairly hot it stayed around 88c to 92c on the cores but I did not get that much thermal throttling. All core clock speed was 4.7GHz.

 

its great points!
i have with default settings (power limit and voltage offset 50) only 1800 points and always overheating (more than 99C), frequency drop to 3,6 ghz, then temp are 80C
What bios do you have - Prema, or unlocked manufactory or locked stock?

What Stepping show CPU-Z program? P0 or R0

 

I don’t have my P775 on me at the moment I just left the house, but when I get home I will check the stepping for you CPU-Z

I have Prema BIOS with CPU at stock settings with -120 under volt. That sounds super hot to me especially throttling at 3.6ghz. What thermal paste and thermal pads did you use?

 

CPU are delid? or you dont know?
part of what you are have great performance and temp - it is bios from Prema! i dont have his bios but have all unlocked options for change performance...
my thermal paste is arctic cooling mx4 and GC Gelid Extreme. (no difference, maybe 1 or 2 degrees but its dosnt matter becaus cpu is very hot). and i have single CPU heatsink and single GPU heatsink. its worse than double (common heatsink)

 

I have no idea, I can ask my supplier and find out for you. But I dont think it is.

Hmm that is really good thermal paste and pads but yeah this CPU does get hot lol

I have the heatsink that has the shared heatpipe so it is one big heat sink not indivdual. Maybe that is why?

 

i think maybe but its part of averall (+prema Bios + best Cpu revision abd best undervolting)...

 

Hi guys, just giviing you all an update. I was able to almost remove my coil whine so it is not audible.

 

The absolute value matters too as chips can ship with different voltages stock.

 

Oh really I did not know that I am a noob still. I will get the voltages when I get home. What is the best way to get readings for them?

 

That's actually a little tricky if you want to do a chip to chip comparison properly. You can see the requested value under load in programs like hwinfo.

 

I ran Cinebench R15 for 30 times in a row, I ran Time Spy test and Fire Strike test twice and I played COD:MW and these are the voltages I got from HWMonitor.

The lowest VID min value was 0.510V and the highest VID max value was 1.379V, I hope this helps

 

By the way, the problem with my 1080 was again dust, dust, dust. After cleaning the fans, it keeps a cool 65°C. Not that anyone asked

 

Always worth cleaning out every so often.

 

[QUOTE = "thecodelifter, post: 10967488, member: 725498"] I ran the Cinebench R15 30 times in a row, ran the Time Spy test and the Fire Strike test twice, and played COD: MW, which is the voltage I got from HWMonitor.

The smallest minimum VID value was 0.510 V, and the highest maximum VID value was 1.379 V, I hope this helps [/ QUOTE]
Thanks for test. my voltage with -50 undervolts in HWinfo - hotest core have 1.41v

 

The best way to compare is to look at graphs under similar loads.

 

So I've got a pair of 8GB RAM modules from my old NP9870-S that got killed by some rough handling a few months back. The display got trashed and the case was broken but it still worked fine when hooked up to an external monitor. The order for that one said the RAM was "Kingston HyperX CL14 Dual Channel DDR4 SDRAM at 2400MHz". My current NP9176-G3 order breakdown says it's got "Dual Channel DDR4 SDRAM at 2400MHz".

Is there any reason why I can't salvage the RAM out of the old laptop and put them in the new one to bump me up to 32GB of RAM? AFAIK, the only thing I need to do is match speeds, but I've never tried this before.

 

Should work fine, the timings will default to the slowest set.

 

Cool, thanks. Next question: Where's the empty RAM slots on this unit and how do I get to them? I want to do the minimum possible teardown so I don't want to go on a scavenger hunt for them, heh.

 

Indeed, cleanliness is next to godliness!

Just gave mine a once over after installing the custom ihs. I was surprised at how much gunk had amounted behind the CPU fan and on the tips of the blades, only after a couple months.

 

I've tried undervolting this 1080 and it acts really weird as if it doesnt stay on the setting. evenyually creeps back up to 90c-90c+ I believe I'm on or around 70-80watts with undervolt on cpu as well. The cpu seems manageable. 4.5 all six cores at around 70c its this 1080. also with msi curve I can undervolt like 0.825 with below 90c sometimes but then very low clocks as it doesnt like overclocking after the undervolt like my old 1070 did. my 4th gen gt72 is undervolted OC 0.993 200 on core 500 mem at like 67c and thats with cpu maxed out usage at 3.4 all 4 cores at 70c with a 100.6 undervolt. maybe I just have a bad card or I need re-pasting

 

oh and dx11 and future rendering on

 

Under the keyboard if the main hatch ones are not populated. Screws are marked on the bottom and then its help down by strong magnets and clips.

 

Hi!
I have an interesting issue with my P775Tm1-G.
I have latest version of BIOS, Win10 and NVidia driver installed and I have an interesting problem.
If I cold boot the laptop everything is working fine, however if I restart Windows after it boots up G-Sync does not work correctly. Some games run fine, however many start to stutter and Fire Strike graphics score drops from 24000 to 14000 points. Only way to fix this is to avoid rebooting but even if I accidentally reboot windows, I need to shut it down and start up the machine again and when I do that everything works fine again.

Anyone experienced something like this?
Is this a driver, windows or bios issue ?
Why does it work fine if I start it up and what breaks if I reboot it instead of shutting down ?

 

What happens if you do a display driver reset with windows key + ctrl + shift + b?

 

Nothing. I mean the screen flashes once, but FS performance does not drop like when I do a Windows restart.

 

maybe need reinstall CCC and NV driver.

 

I have some driver issues too here and there. If I use MSI AB my system runs 3D clocks 24/7. If I don’t use msi AB it’ll downclock to normal 2D levels. But then I can’t undervolt my GPU, or overclock my GDDR5X memory.

Also sometimes when I launch a game it will only boost to 398Mhz lol and it shows 100% GPU utilization. Closing and re-opening the game usually fixes this.

I just deal with it. If I want to overclock or undervolt my GPU it’ll never downclock to 2D levels because I require MSI AB.


I’ve reinstalled windows, ran DDU, changed drivers for the GPU. I even changed the settings in MSI AB trying to fix it with absolutely no luck.

Another thing too, my Clevo GTX1080 is actually a factory overclocked model, so it’ll run 1,647Mhz with 1,847Mhz boost by default and even downclock to 2D low power speeds. But, once I run MSI Afterburner it switches from 1,647mhz default to 1,556Mhz default and forces 3D clock speeds all the time.

It is very stupid. And I wish there was a fix for this. I wish I could just render a overclock and have 2D clock speeds at the same time.

 

Ccc is not installed, using Obsidian software.

 

Unlikely to be a display driver issue then.

 

After carefully testing I can confirm it is.
Installed standard Nvidia 441.20 and 441.34 with ddu clean install and both had the same issue which the previous 3 Nvidia drivers had:
Fire Strike graphics was normal and laptop Gsync worked fine when laptop cold booted, however testing same things after reboot I could see stuttering and FS graphics score dropped from 24000 to 14000 and remained like this until I cold booted again.
Tried the 5 latest standard drivers, desktop and laptop and all had this bug.

What I did to fix this is that I ddu unstalled Nvidia driver and installed clevo 431.40 dch driver. After that I tested the laptop after reboot and performance was fine, no stuttering or fs gs drop as seen with standard 441.20 Nvidia driver.
Just for the sake of science, I did not uninstall clevo 431.40 and instead express installed Nvidia 441.20 dch over it keeping the settings.
I rebooted, cold booted did everything which triggered the bug before, but did not happen again, performance is fine now and with 441.20 Gsync is slightly smoother than with 431.40.

I dont know if express installing over clevo 431.40 dch driver or using dch instead of standard fixed it, but it is working fine now.

 

Now all what remains is to find a fix for the 8700k microcode issue found both in .25 and .27 clevo BIOS .

 

What issue is this?

 

441.40 driver fixed my constant 3D clock speeds. My GTX1080 steps down to 2D levels now.