Clevo P775TM1-G questions

I just bought a P775TM1-G with a 1080p g sync display. RTX 2080 and i7-9700k. Do i need to swap the motherboard in order to swap in a B173QTN01.0 17.3 inch Screen 3K 120HZ LCD Screen 2560x1440 EDP 40 PIN Non-Touch display?

I really prefer to have a 1440p screen.

 

Congratulations,
No need to do anything with your motherboard, only the right screen and the right cable. And you can find all details about compatible screens in your service manual. And you can ask your seller.

 

I bought this from Pro Star computers but no one really speaks english well so was easier to come here to ask. What should i expect to pay for a 1440p screen? Will i lose G Sync? Is my brand new 1080 144hz G Sync screen worth reselling?

 

Your screen is fine and you can sell it without problem, or you can just keep it as a backup spare part. You can use the 4K one if you need it.

@Prostar Computer can you please give some insight.

 

Not planning to go 4k as in my opinion it isnt worth the low refresh rate. the 1440p screen seems like the sweet spot right now. Prostar doesnt offer a 1440p screen so i have to go aftermarket sadly

 

This is your service manual from @Prema website https://premamod.wordpress.com/:
Download link:
https://mega.nz/#!rZ8B2TjS!qNtFi2Je8SfSZuxnaR8MKaN1OgNEMSGar8eNuYh0yXo

 

They need to update it, it doesnt include current gen hardware.

 

This Sansung 970 500GB is way too hot, right?

What can I do to cool it?

Thanks.

 

Look for any SSD Low-Profile Heat Sink on eBay or Amazon, and make sure that you are using the updated firmware from Sumsung.
https://www.amazon.com/Heatsinks-Laptop-Cooling-Radiation-Dissipate/dp/B01MY2NQ8S

 

In addition to the heatsink, make sure you install it in the M.2 slot under the GPU fan as this area gets more airflow.

 

Depending on your use case and the amount of airflow over the SSD, you could lower temps just by adding a small heatsink to the SSD controller. I "glued" one of these heatsinks to the controller on my 960 Pro using Arctic MX-4 and the temperature spikes decreased noticeably, even more so when using the cooling pad. The main reason I went for this approach is this GN video. In my case, sinking the memory would likely not be beneficial.

 

Yes, using the latest firmware.

I have 2 x 500GB 970 EVO's in RAID0.

I was installing Metro Exodus at the point I took the screenshot.

I'll have a look at a small heatsink.

Thanks.

 

A thermal pad to the case or very slim heatsink. Too big and you will increase thermal capacitance but reduce cooling.

 

my 960 evo also hits 95C at times. i should do this?

 

You don't have to but you can.

 

Hello guys!
I've just bought a Clevo P775TM1-g with RTX 2070 / i7-9700k / 8+8 Gb 3000 MHz RAM / PREMA bios and i'm looking for some hints on how to configure it.

But at first a bit of history: i've bought it several weeks ago and at the beginning it was with bugged EC (EC always seen the GPUs temperatures as 40C and fan on the GPU always been at 20% - there is some thread here on notebookreview ( http://forum.notebookreview.com/threads/prema-bios-gpu-fan-problem.827527/) with guy that had the same problem as me) and after week seller had given me an updated BIOS that solved this EC problem. So i've deleted control center, updated BIOS and tried to run a few tests - in Furmark it was OK after update same as in SPECviewperf13 and Battlefield V - fans have been working just fine stabilizing GPU and CPU temperatures at ~70 - 75 degrees celsius. (here is the SPECviewperf13 results with Afterburner log http://4pda.ru/forum/dl/post/15606406/SPECviewperf13_RTX2070_afterburner.zip, pity but i have logs only for SPECviewperf13). Finally, pleased with the results i'd decided to install Control Center (seller proposed me to install CC from thei site) - and in Furmark all seemed ok but when i've run Battlefield V i've meet up to 95 degrees CPU temperatures. I've tried to uninstall Control Center and re-update BIOS - but the situation have not changed (here is the afterburner log http://4pda.ru/forum/dl/post/15657765/HardwareMonitoring.zip - first log for deleted CC and admin account, second for deleted CC non-admin account, third for installed CC and admin account and the fourth is for installed CC and non-admin account).

And currently i'm not sure what just have happened - possibly some power limits/voltages/frequency multiplier or factor have been shifted when i have installed CC at first after BIOS updating. So my main question is - is it normal temperatures for this laptop and this processor (CPU frequency/load/temperature you can observe in afterburner logs - currently it configured to limit it's multipliers to 46 when all cores are loaded (and it works on 4600 MHz when all cores are loaded)).

And my next question would be "how should i configure my laptop? What limits \ frequency factor or multipliers should i specify" for better work?" It is ok for me to slightly sacrifice by the performance to obtain better comfort and to extend laptops life.

Thank you very much for your answers!!!

(here i've uploaded files on other server https://gofile.io/?c=37z5l4 - 4pda.ru requires sign-in to be able to download)

 

Battlefield is a heavy title on both CPU and gpu. You can undervolt to see how much power saving you can get on the CPU.

 

By the way, here is the result of LINPACK running on OCCT - it is turning off due to high temperatures. What should i change in configuration? Suppose, i should change powerlimits or frequency multipliers somehow...

And here i've run CPU-z bench - ~90-92 degrees is not ok for me, what should i do?

 

Well, i've checked again afterburner logs and found that earlier when i'd played battlefield V CPU consumed ~80-90 W (working on 4600 MHz) and now i've tried to play again and it consumes only 40 W (working on 3600 MHz - nominal) and the temperatures are again around 70 degrees. Something had loaded my CPU previous time (possibly some Windows updates ) or now for some reason it dont want to boost, but nevermind, now i see that in some cases my CPU could consume lot of power and go to dangerous temperatures. In Control Center there are such PowerLimit parameters now: Power Limit 1 = 120 W, Power Limit 2 = 165 W, "time allowed to be in extreme" = 26 s. Should i change smth?

Thank you again!!

 

From my own testing and stuff, i would first try testing a voltage offset on the CPU, something like -80mv. But i wouldn't hope to have temps bellow 92C. I've been having temp issues since i got the laptop. BFV is a good game for stability testing.

 

So, is it normal to obtain such temperatures with these specs?
Are thiose power limits are good? Should i change them?
Should i decrease frequency multipliers/factors (for example 43-44 instead of 46 for all cores)?

Does it really dangerous to have such temperatures? I'm planning to use this laptop i the next five years, at least.

 

Well by right we should be able to run @ 4.7 all cores but we cant. I've been running my CPU at 4.4 with 92-94C for a while in games like BFV because nothing i do does anything to the temps, ive just about given up.

 

By the way, how to explain the fact that yerstaday my CPU on BF5 had been working on 4600 on all cores and today on 3600 with the same performance actually (playing on ultra + DXR + 1080p ~ 60-70 fps)

 

Check your BIOS core ratio setting?

To be honest I'd wait for someone with more knowledge then me to reply to that.

 

The best way to know what your system is doing starts with graphs and ends with graphs with some extra graphs in the middle. CPU usage, GPU usage, temps and clocks for each with throttle reasons. Mapped over a few gameplay sessions.

 

Can’t compare your model vs. same model with Turing graphics. You have different heatsink and 1080 graphics with higher TGP. Op should undervolt.

 

Which tests should i provide?

(BTW, seems that it is normal for my system to heat so much - so, possibly i should unvervolt and lower a bit frequency)

Will my CPU and laptop in total degrade faster with these (~90 degrees) temepratures than with some 70-80 degrees for example?

 

Physics says yes, the hotter the faster normal degradation.

 

Here i've made few tests - system state has been captured by MSI afterburner and saved in one log (all tests per one session):
http://forum.notebookreview.com/att...6/?temp_hash=73e4a3f6da89f79a7ecb7a8674e7a924

stress в CPUz:


furmark:


stress в CPUz + furmark (very hot):

 

BF5:

and here what i've got on HardwareMonitor after game closing (i've rerun it before the game but forgot to make a screenshot during the game so only MAX values could be usefull).


Well, this time during BF5 CPU worked almost all the time on 4600 MHz (sometime falling to 4400 that can be seen from afterburner logs) and not falling to defolt 3600 MHz as previous time. And that strange if be honest

Nevermind, what can you say about my system? Is it work properly?
(As far as i understood it works just as it should - CPU is really hot but in the same time it provides good performance and so i should find some balance. I've seen some Polish guy on youtube that has nearly same specs / temperatures)

 

I understand that, but even then, his CPU is putting out less heat then my 8700K with less cores and clocks. It doesn't make sense.

 

Also thing to note BFV uses AVX so if you have AVX offset it will downclock your CPU. BFV puts a lot of stress and its a good way to test your system and your tweaks.

 

Less cores? I expect you meant less threads. 9700K is 8 cores and with higher boost clocks at stock vs. 8700K. And he still run without undervolt. I expext you run 6 cores [email protected] as pict below. Delidded W/ Liquid metal. If @AntonyCooper run stock clocks he should undervolt. Same also if he put a slight overclock. There is no need to use stock voltage as it is.

 

I mean my CPU has less cores with an undervolt, lower clocks and delidded yet i'm still running with higher temps. That does not make sense to me. I've done almost everything to make temps better but to no success.

 

Fine, in order to make an undervolt should i remove CC, install intel XTU and do the optimization or i can left CC installed? Some guys have problems with CC and XTU profile - CC overwrites XTU configs or smth. I am really new to this and have never made an optimization like this - so if there are some tricks or hints could you share it with me, please ?

(About temperatures - my CPU has less threads, soldered, and pasted with Grizly Kryonaut + i've run BF5 with ultra preset -> GPU near 100% load when CPU is nearly 70% load, if i'll downgrade graphic and increase FPS CPU will be loaded more for sure and temps will go up)

 

Maybe you miss understood my post. Your model with 1080 don’t have Unified Heatsink. The refreshed model with Turing graphics got the new Unified heatsink(one heatsink for CPU and GPU). This type heatsink can cool down the processors better if you run mostly Cpu load. + RTX 2080 graphics have lower TGP vs. GTX 1080. Clevo gave the Cpu better cooling due crippled TGP on Turing graphics.

 

BTW, what do you think about "LENOVO Y GAMING ARMORED" (a backpack for laptop).

And is it reasonable to use some "screensaver mouse pad" - microfiber that is placed between keyboard and screen to prevent screen damage (on my previous laptop with time and carrying it in backpack i've obtained spots on the screen with higher backlight, dont know how to explain - like scuff in form of path of these letters on keyboards "qwertyhbvcxz" like letter U rotated on 90 degrees). I am not going to take laptop outside of my apartment too frequently but from time to time i'll bring it from one city to another and so i'm trying to find a safe way to do this. (Currently i have Everki Commute 18.4 - but it useful only for some short trips or as сase for storing a laptop at home)

 

I've watched video about GPU undervolt - OwnOrDisown channel, where the guy optimizied frequency/voltage curve for GPU but as for CPU i have not found any such procedure instructions - only some constant offset. Seems that i have no control on frequency/voltage as in case of GPU but only on some parameters of the predefined curve (could shift this curve donw for example, and possibly there are few other parameters). Am i correct? Or there is more accurate way to optimize CPU (I've seen one article on russian source (i am russian native speaker so it's not a problem for me) habrahabr but have not read it properly and seems that it could be not applicable in my case)?

UPD. Lurking about ThrottleStop - there should be more accurate instrument for optimization.

 

I get that, but still doesn't explain why my temps are higher then others with the same identical config. 8700k / 9900k with gtx 1080.

 

Are you sure that you've compared temperatures for the same "CPU load" (or GPU load)?

 

CPUs have a finer curve normally and it fits the core pretty well.

 

Yeah, but when i'm applying some voltage offset - it would act same for all frequencies, right? CPU voltage during it's work is changing with frequencies changing: for example, lets say that it will operate with 1.2 V on 3600 MHz and 1.4 V on 4600 MHz, then i'm applying some 0.1 V voltage offset, after that it will operate with 1.1 V on 3600MHz and 1.3 V on 4600 MHz, and for 1.4 V drop on 0.1 V means ~7% drop but for 1.2 V drop on 0.1 means ~8% that could lead to different effects i suppose.

But possibly, my understanding is not correct, possibly this voltage offset is applying with some scalling?!

Nevermind, here is the link to article https://habr.com/ru/post/228233/ (on russian) where guy has optimized this voltage/frequency curve for some old CPU.

 

No scaling though the margins are pretty similar along the curve these days.

 

Ok, what software better then to use? Intel XTU or throttlestop (i am not sure, but on some forum some guy said that XTU conflicts with Clevo Control Center, is it true)?

BTW, i have PREMA bios so i can specify Vcore offset just in BIOS or in CCC (there is also offset preference).

 

I would do any overclock or voltage offsets in the BIOS, don't use the software.

Try offset of -90mv adaptive offset and 4.7ghz and play a round of BFV if it's stable and temps look good, try more offset.

 

Currently i've made -0.13 offset + set the powerlimit1 to 95 W. Looks better - CPUz + Furmark now heat up the CPU to ~85 W instead of 115 W (and 83 degrees instead of 95 degrees), in BF5 also better temperatures (~70-75 degrees, but CPU is not fully loaded cause i've set ULTRA preset with DXR - GPU 100% loaded)

(and yeah, i've removed XTU)

Seems that it is possible to set even bigger offset and it will work with higher load (when CPU is loaded and it's voltage is higher then 1 V) but when load is small (and voltage drops below 1V) it crashes - if there was a possibility to provide and "adaptive offset - specify higher offset for higher frequencies and lower offset for lower frequencies" i would be great

 

Why do you play BFV with DXR? I don't think for a FPS that DRX is really that needed but thats just me.

You should be able to have -90mv or more from what I've seen. You just have to do some testing to see what your CPU is able to do.

 

Sure sure - actually i've played only to test GPU before and have not reconfigure it ) - i will test it with lower graphic configs later, but i've conducted stress tests in CPUz and OCCT (simple + LINPACK) and all is ok with this -0.13 V offset.

-0.13 V is -130 mV that is already higher than your -90 mV. But as i said, it is possible to go even further and decrease voltage more.

P.S. About DXR - where else i will see DXR work with mirrors and all these staff?! (I've played a bit in METRO - but there DXR is used for other thing)

 

That's great if you're stable with those undervolt. Looks like you have a good CPU.

For testing and seeing how it works I would turn DXR on for a few rounds but after that I'd rather play with smooth 140 fps then 70fps.

 

It's too expensive to use everywhere at the moment, future games will be able to start using it more often as the RTX speeds increase.