*** The Official MSI GS65 Stealth Owners and Discussions Lounge ***

At the risk of not knowing exactly what the state of the situation is, as soon as you find a serious fault in a new laptop, ask for a return and cross shipment swap for a new unit. There's no need to pussy foot around, just get a new one.

You may think ordering a new fan will solve the problem, and it might, but each day that goes by reduces your "return without question" period, usually 7-14 days, so don't let that time expire - you could be stuck needing an RMA back to the maker to repair it.

And, the problem might not be the fan itself, it might be another problem on the motherboard, a dead sensor, a bad ribbon cable, etc. So rather than debug this remotely, send it back and get a new one.

 

We're currently assisting him, thank you hmscott.

 

Hey there guys. Long time no see.

Recently, I've pulled the trigger on a 1 month old, GTX1070MQ MSI GS65 with 32 GBs of RAM and with an extra 256 GB Samsung EVO SSD plugged in. Checked everything as i could imagine before buying it and everything was perfect. The price was lower than the basic 16 GB model. (it has a 1070 max-q). After that, I've ditched the AW 17R4. Boy what a relief to own a 4 pound laptop this thin after that behemoth. And as for the flex goes, I was relieved beyond words that the flexing is sooooo exaggerated. It is much better than my old Clevo notebook, and i thought that notebook was sturdy. I couldn't bend the area around the power button, i don't know what people are talking about. There was no creaking sounds also. Only that the laptop is making some noises while heating up - cooling down, thats it.

The temps were terrible out of the box, hitting 94 degrees and throttling around %30 on AIDA64 burn test on a custom fan curve. The laptop is on stock paste BTW.

First thing i did was to create a custom fan curve fans maxing out at %120.
Then I've managed to undervolt the CPU -160mV (stable) on x36 x35 x34 boost clocks, and i've managed to contain the temps under 80 while playing Assassin's Creed Origins which is perfect for me. it occassionally reaches 82-83 but i think that is fine. no throttling whatsoever. And for the gtx 1070, i've managed to overclock it with the boost clocks of +150mhz core and +300mhz for the memory. +200 on core doens't seem stable for me, crashes both BF1 and ACO. Before downclocking the CPU to 36-35-34, i was reaching 13.500 scores in firestrike. After the final setup, i'm getting around 12500-13000, not sure, gonna check it again. The highest temp i've seen for the GPU is 76. It keeps the clocks around 1500 - 1700 most of the time while playing ACO but sometimes downclocks to 1100ish clocks on normal 70ish temps. Is it throttling itself down? Whats the throttling temp for the max q 1070 on this thing?

AIDA64 is still stressing the CPU all the way up to 90 degrees but i don't think that test is relevant for me at this point.

Still using MSI Dragon Center, applying the GPU overclock before getting in to a game since i think the Dragon Center resets the Overclock.

I don't think that the -160 mV undervolt will hold other than these downclocked operating frequencies because if i raise the clocks up to 39 38 37 it crashes on BF1 and ACO. Some more testing is required i guess.

I'm open to suggestions for tweaking this baby beast. I think i've yet to find a sweetspot between a minimal performance sacrifice / heat trade off.

note: the only downside about the build quality i could observe is the long plastic part between the hinges, which goes from one hinge to all the way to the other one. That part makes some little noises if i push a finger on it, and doesn't match up with the build quality of the rest of the laptop. so thats that.

Final note: I live in Turkey, so I don't have any access to some resellers who provides any repasting services. This thing costs 3260 dollars here. No kidding. We are buying one for ourselves and one for the government everytime we purchase stuff. I could have built a NASA computer with this kind of money. SO when its this expensive, I don't have the guts to open this thing and make a repaste on my own *yet* but when the warranty expires, I think I'm gonna go for a repaste. Because my honest opinion is that we pay for the "3.9 GHZ on all cores" performance equipment, and thats what I will try to achieve in the future. I honestly think that a repaste is essential for this machine. But at the time being, LETS THROTTLE OUR MACHINES BEFOREHAND IN ORDER TO PREVENT THROTTLING !

 

Anytime I start a game the cpu jumps straight to 89c-90c, there's no build up to that temp. That's the max temp, but it comes instantly. Is that normal for this laptop?

 

No. RMA it. Don't even bother trying to fix that. Just RMA. Trust me.
Just. RMA. It.

 

Does anyone know of a case or sleeve that would fit this laptop? It’s too big for 14 inch and too small for 15.6 inch sleeves

 

Just get the 15.6" inch sleeve. You don't need to be a perfectionalist.

 

So that's not typical? I've been reading other people's comments about it getting to 90c. Does it take a lot longer or something?

 

I had one custom made at 1upkeyboards.com. Just pick a color and give them the dimensions of the laptop. Fits perfectly and very good quality.

 

Mine gets that hot when I am under extreme load. The physics test in Firestrike for example. If it's hitting 90 just booting, there's a problem and you're just going to be throttling like crazy.

 

My firestrike physics test also hits 90C right away, but I've seen others who can get through it without thermal throttling on stock paste

 

Run a full FS test with HWinfo and send a screenshot of your temp graph. Core 2 and core max. That way we can compare.

 

Here are some plots of the HWinfo logs. I had coolerboost running the whole time. The first one is stock paste, stock voltage. The second is stock paste, -230/-130 core/cache, turbo multiplier 39. The undervolt helps a lot, it reduces temps by 10C, but I still hit 90+ as soon as the physics portion of the test starts. I think I could maybe hit not throttling temps with an aftermarket thermal repaste

https://imgur.com/a/1tKwUVg

 

That seems excessively high for being UV and with cooler boost. Others have reported mid to low 80s.

What's the CPU throttling to on the first test without the UV?

 

Post the HWINFO screenshot of the maximum CPU VID please.
If it's higher than 1.25v with that huge undervolt then something is 100% purely wrong.

 

Of course you se less throttling. You run the processor below it’s default clocks. Same as I run old 6700Hq 2,6GHz instead of default 3.1GHz.

 

I also forgot to mention my ambient temps were only 13-15 C when I ran these (cold canadian morning). It throttles to 3.4 GHz and trending downwards by the end of the test

Here is the screenshot and my voltages during the test. I'm very happy with my undervolt (haven't seen anyone else stable this low), but I definitely agree there is something wrong with my temps. I just hope a repaste will fix it and it isn't the silicon.

https://imgur.com/a/jUwf65o

 

I don't think anyone has throttled that bad yet. I would RMA before repasting.

 

Yes, what i meant to say was, i've done my undervolting according to the downclocked multipliers...

 

VID of only 1.05v and reaching 90C?
Please RMA. You have a pretty serious problem. Looks to me like the heatsink is not even touching the die.
I could run without any thermal paste at all at 3.4 ghz and 1.05v and not reach 90C in 5 seconds...

 

@Donald@HIDevolution or @Phoenix

I want to dual boot Linux. What distro/s will work best and how would I go about doing this/specifics?

Planning to get the gs65 1060 version.

 

Try this tutorial.
https://itsfoss.com/install-ubuntu-1404-dual-boot-mode-windows-8-81-uefi/

 

I have encountered something weird. Nahimic services are causing uTorrent to hang and become inresponsive whenever i launch it. If i terminate nahimic services from the task manager, it immediately starts to operate normally. I uninstalled nahimic, and the problem is gone. Guess uTorrents media player capabilites are conflicting with the nahimic ones.

 

Hello everyone. I am still having some problems with my undervolting, have had some blue screens of deaths, but nothing serious. I haven't actually changed any of my undervolting settings through TS, but my powerplans on Windows might be causing this problem. Does anyone know what the minimum processor state must be in the powerplans, because i think it might have something to do with that (BTW., i have Speedstep enabled in BIOS)

 

It's probably only the spike caused by BIOS setting AC loadline at 180 for the default that's why the HWInfo captured 1.05V, I played around with various setting on the AC loadline, setting to 1 will make the voltage stay unless the multiplier has changed, but setting it to 180 I see the voltage fluctuates a lot even there's no change in the Multiplier. AC loadline seems to control the range that the voltage fluctuates for each Multiplier / Cache ratio.

Currently with AC loadline set to 1, my CPU core is set to -165mV (But I feel that when core is undervolted to a certain point, it won't lower further) but CPU cache is set to -43mV, and the voltage for each multiplier is as followings:

x32 = 0.8948 V / Cache ratio = 29
x33 = 0.9148 V / Cache ratio = 30
x34 = 0.9348 V / Cache ratio = 31
x35 = 0.9548 V / Cache ratio = 32
x36 = 0.9797 V / Cache ratio = 33
x37 = 1.0046 V / Cache ratio = 34
x38 = 1.0247 V / Cache ratio = 35
x39 = 1.0497 V / Cache ratio = 36

The reason my cache is only to -43 mV was because if I go for more I get some error count when benching with ThrottleStop's built in bench, I use this to test if the memory or anything gets unstable because of too low voltage and reports error.

The AC loadline greatly affects the undervolt setting you can go, for examble, in default setting which is 180, I could undervolt my cache past 160 mV, but with AC loadline set to 1, max I could go is -43 mV before an error appears, but both settings have the temperature performed similarly.

You were right Falkentyne, with the guide you posted before, you won't be able to undervolt much, but I do prefer the way you posted, the voltage is super stable and does not fluctuate like the default setting.

And by the way I tried some more tweaking on my RAM I do see improvement with my ns in AIDA64.

This is with 1.20V 2400MHz 15-15-35-400


This is with 1.25V 2666MHz 18-18-40-430


This is with 1.25V 2666MHz 16-16-38-420


For some reason if I lower the RFC from default 420 when in 2666MHz, the laptop will not boot, had to do CMOS clear for it to reset, so I didn't try your setting for the RFC.

 

Has anyone experienced where the sound card suddenly stopped working? Speakers nor audio input or output work... Windows doesnt recognize its Realtek, even though i have re-installed the drivers an all possible ways. Help!

 

This is EXACTLY why I posted the Bios unlock code left alt, right control, right shift, F2.
You see proof of the results yourself.
FULL STABILITY with NO drawbacks and NO voltage spikes with AC DC=1.
The only problem is the high temps and that's MSI bad heatsink QA lottery design issues we can't fix.

The spikes you see at full idle to very slight load (esp after windows boots) with AC 180, is the actual VID boost range as you said. This is the maximum VID boost that will occur at MAXIMUM LOAD---e.g. if you tried prime95 small FFT (AVX disabled with CPUSupportsAVX=0 in the local.txt folder in prime95 folder). That's the actual maximum voltage the chip will get.

Unfortunately, due to a DESIGN ..."SHENANIGAN" with these MSI laptops, the VID you will see with that prime95 test, with IA AC=180, DC=180, is **NOT** the actual maximum VID! The reason is because of the IA DC value itself !! The DC value controls how much the VOLTAGE will "drop" at maximum load, and then reports that as VID. The AC value controls how much the voltage will RISE at maximum load, and the VID baseline is there.

Let that sink in a moment.
There is no VCORE monitoring sensor in these laptops.
The CPU VID is what voltage the CPU is requesting or 'thinks' its getting.
But the CPU actual vcore its usiing and VID its reporting can be very different.

Basically, the AC setting controls the voltage going into the CPU and the voltage regulator will boost that at higher loads.
The DC setting controls what "VID" the CPU reports (think of it as what is going out of the CPU) and will "droop" by a factor of 1.80 mOhms (the higher the current, the higher the resistance, thus the more vdroop).

But the VID is actually lower than the true voltage at full load with IA AC =180 DC=180, and is mostly accurate with IA AC=1 and IA DC=1.

The reason for this is some hidden setting MSI uses, called "Loadline Calibration" which removes the vdroop that the IA DC setting is supposed to be reporting..
However this Loadline Calibration functions on the vcore signal itself and counters the vdroop, and the IA DC setting cannot read or understand this type of setting, as intel cpu's are designed around vdroop being active at full load. So that is why the CPU VID shown at full sustained load with AC DC=180 will be much lower than the actual real voltage. You can see this by temps and power draw (watts) being reported in a clearly strange way.

If you want to see how much vdroop your laptops are supposed to have by default, set IA AC =1 and IA DC=180
This will not change stability.
Then monitor the VID in HWinfo64 or Throttlestop 8.60 at full idle.
Then put a heavy sustained load on your CPU.

Watch what happens to the VID. Prepare to have your mind blown.
The VID will drop substantially but the actual voltage (handled by the IA AC setting) will remain almost the same (you can tell by the temps and power draw).

If you want to see how much the voltage truly rises at full load, set IA AC loadline=180 and set IA DC loadline=1. Then the DC will not droop at all and the full voltage spike will be in effect.

 

I run Ubuntu 18.04 for work. Also I did try Linuxmint 19. Both distros are running fine.

 

With only the custom fan curve, and without any undervolt or downclocking, my CPU also reaches to 90 degrees 'till i reach the actual gameplay after pressing PLAY on Assassins Creed Origins. When the loading screen appears with only Bayek on the screen, it reaches about 82 and when i start actually playing, it is already 89 - 90 degrees.

Note: Didn't see the UV part sorry.

 

Can you launch dragon center and inspect the fan 1 & fan 2 speeds?

 

I've used one of those blank profiles.

It's shift is on "Turbo"

The CPU fan curve is advanced, with %30 - 50 - 70 - 80 - 100 - 140.

Their RPM's are: 6000 fan 1 - 4800 fan 2

But this time, GPU usage varied around %50 and the temp was 69. CPU usage was %40 - %50 and the temp was 80 after 1-2 minutes of gameplay. I think i was in a very heavy area the last time i've done the test not sure.

If i switch to cooler boost mode: fan RPM's are 6000 fan 1 - 6400 fan 2,
and when the GPU usage is %90, GPU temp is 70.
CPU Usage %40, CPU Temp varies between 75-80 after 5 minutes of gameplay.

CPU clock is always at 3.9 Ghz.

ThrottleStop has not been launched at all after restart. (So that means no undervolt right?)

So yeah, ozbdfdqj might have a problem with his device.

 

Ok I will try those thank you.

 

Thanks I’ll look into this.

 

Yes I am experiencing this issue and it's more common when I am on YouTube. I didn't find a solution yet.
I was wondering if there is an app to update all the MSI hardware drivers on Windows 10... @Phoenix

 

Relased today a new BIOS


New BIOS : E16Q2IMS.10C
ROM CheckSum : 0x7B5D
Release Date : 2018/07/06

;--------------------------- Description ------------------------------------;

1. Improve the stability of the panel display.

 

I've experienced this when I was on E16Q2IMS.105, after updating BIOS to E16Q2IMS.10A I don't see this problem anymore.

To get it fixed, you'll need to do CMOS clear, it worked for me every time when I was on E16Q2IMS.105:

1. Unplug power cable
   
2. Hold down the Power button for 30 sec ~ 1 min, at this time if you don't have the power cable plugged in you should not be able to turn on the laptop.
   
3. Plug in the power cable and turn on the laptop, it should be fixed.

Please note that doing this action will reset the date in the BIOS back to 01/01/2015, but the BIOS will obtain the newest time and date once it boots into Windows, so don't need to go into the BIOS to set the time and date again.

 

I see, thanks for the thorough explain. Glad to know that I'm seeing the real voltage reported by the CPU with this tweak.

 

By the way, we will never be able to see the full version of change log but minimized/reduced version of it. This is why it's always recommended to go with the latest and greatest if we can. If it's released, it usually made for the better than worse.

Here's an example of a full version log.
Code Base:4.6.5.3_VEB_1AQPM037
ME FW version:Intel Management Engine Firmware Release x.x.xx.xxxx
UMA Vbios version:2175
GOP driver version:5.0.1032
PXE OpROM version:Bigfoot-pxe-v2.1.1.5
UEFI PXE driver:LxUndiDxe-X64-v1.1.0.4
Verb Table version:HDACfg-0892-xxxx10EC_201x0123_Realtek LOGO
UEFI Raid Driver : 1x.7.0.1936
Raid Oprom : 1x.7.0.1936
Bios Source:ExxxxIxx20xxxxxx_CRB37.RAR
AMI VEB Tools Label: 4.5.5_TOOLS_40
Compiler Tool(s): Microsoft¡¦s Windows Driver Kit (WDK) Ver. 7600.16385.1
;------ Description ------;
;---------------------------------------------------
1.Add "Intel Virtualization Technology" option on BSU.
2.Add the info of "Intel VT-x Technology" on BSU.
3.Modify GC mode list to fix win7 logo issue on UEFI with CSM mode.
4.Modify Ctrl + Home recovery BIOS behavior.

 

Have you had any issues with the wifi adapter not being present after waking from suspend?

 

Thanks Kevin. I may wait for others to test this first. My system is running cool and quiet. I want to find out if they made any changes to the fan profile which cause it to run louder

 

Did anyone notice that the temperature on the CPU gets like 5-8C higher when benching both the CPU and GPU than benching only the CPU?

I thought the thermal design in this laptop was to clearly separate the heat handling for each side of fan? Was I wrong?

Could it be that a portion of the Heat generated from the GPU gets transferred to the CPU side of the heat dissipation fins through that single thin pipe which greatly affects the effectiveness of heat dissipation from the CPU heat pipe?

Also if you watch the wattage reported by the CPU closely, once the dGPU is active, the wattage reported by the CPU could add another 10W, does activating the dGPU also have any influence to the heat generated from the CPU?

 

On the Titan series, the GPU VRM heatsink is cooled by the CPU fan, so heat is radiated over to the CPU area, thus raising the heat and interfering with the heat transfer. It could also very well just be GPU heatsink heat going in the case and heating up the ambients and raising the CPU as well.

It's one or the other in your case, or a combination of both. But I think the GS65 heatsink is partially unified also, so most likely similar to the Alienware series.

 

Yeah, Stealing cooling capacity from Cpu grills won’t help on the Cpu cooling.

 

Also have to remember that the GPU is going to use far more power than the CPU. It's possible they just needed to do that to help dissipate GPU and it's power delivery component's heat.

And then there's situations like the Tornado F5 where it's best to disconnect that transfer. Lol

 

Hey guys does anyone know which laptop bag/sleeve fits nicely with this laptop?

 

I use this one, spaces are nicely divided. I went working abroad right after I've received the laptop, still working nicely after 2 month and half.

 

FINALLY

I received my Msi GS 65 8 RF 026 IT!
(In Italy, according to MSI, the SKUs number 26 and 83 are exactly the same, the latter was dedicated to megastore but then extended to PC resellers because of a lack of availability).

I hadn't so much time to test it for now, but I was immediately impressed by the smoothness of the screen refresh! Backlight is also quite brilliant for indoor use! I was only a bit concerned about Yellow color rendition with True color profile, which seemed a bit greenish, but I have to investigate.

The PSU is so tiny

The only strange thing I noticed only after receiving the PC is that the ú isn't in its usual place, but moved in the upper row.

Glad to be an active contributor more than only being the thread owner of this laptop.

 

I got mine last June, been using it for the past 18 days. I bought the GS65 8RF-053... The First thing I did was to install an extra ssd (WD Blue 1TB M.2 Sata), and I just found out that the empty slot is for NVMe only, so I opened the laptop again just to swap the two ssd drives. So for those who doesn't know it yet, there it is.

Anyway, about the laptop, this thing is a beast for gaming, it all run my games at the highest settings without a hiccup hehe. the only downside is that it throttles a lot and is very very hot on the side vent areas (especially on the cpu side). It helps a little when you have a laptop stand to have room for air flow or a cooling laptop stand that has lots of space for air flow. The thing is it always throttles but it never affected my gameplay, I don't notice it).

What I notice is that, after gaming, the temps of the WD Blue 1TB M.2 Sata that I installed gets to 65°C but gets back to 45-47°C after awhile. Before gaming, it is at 37°C. Should I worry about it getting to 65°C? My games are installed on the WD Blue drive.

 

Ahoy there, and greetings from India!

My GS65 RF056 just arrived a couple of days ago! It's an absolute treat to use and obviously a huge upgrade from my previous Alienware 13R2, although one thing I'll say in favour of that machine is, it was built like a tank. Don't have any complaints as such with the GS65 build quality - no real creaking, and of course there is minor flex around the power button, but only if I go specifically looking for it, so no biggie.

A couple of questions for fellow owners (the irony of being the Indian guy ASKING tech support questions is not lost on me):
1. Is there any way to program any LED as a disk activity LED? I find it quite surprising there isn't one. Never owned a laptop without it.
2. What kind of battery life are you folks getting out of your machines? I took mine in to work and managed around 4.5 hours with Google Docs on Chrome, some YouTube, and a few heavy downloads/uploads on throughout. This is on the "Better Battery" setting on the battery bar. Any tips for maximising battery life? I'd really like to get 6-7 hours a day of productivity/browsing out of this. Also, does anyone know what the "Power Options" shift in Dragon Centre achieves?
3. Any known dodgy drivers so far? I had some horrible experiences with Dell's Alienware drivers, so a heads-up on any drivers with known issues would be great!

Thanks a bunch!

 

So I got mine a few weeks ago and still can't stop drooling over this beast of a machine. The 4.14lb doesn't hurt either.
I bought a Kingston a1000 960gb to replace the Toshiba it came with. I tried putting it in the first slot and putting the Toshiba in the second slot as the secondary SSD, however, it wasn't being recognized. I pulled it all out and tried again, still not recognized. Consulted my friends, they tried, still didn't work. Switched them, Kingston worked just fine in secondary, and Toshiba worked fine in primary. When I look in the bios though, it says the Toshiba is running on a SATA interface rather than the advertised NVMe, anyone else noticing this? Did I just get screwed and given a SATA SSD? Or does Toshiba just really suck?