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

Perfect exactly what I was looking for, I wont be putting LM I may replace the thermal pads but I will undervolt limit multipliers and have a cooling pad. Still waiting to see how the new rtx 2060 puts out heat, but I imagine it should be similar to the 1070 maxq

 

OK, I wish I could search this thread efficiently and find what other used for the under volt. I've been at it a couple day using XTU and using their stress test to try and get a blue screen or shut down and have not. I've also let computer sit for 15 minutes until screen turns off (I'm assuming this is good enough to check it at idle) and still no problems AND I AM AT -.170!!!

Anyone else go this low? BTW I also ran a 10 minute AIDA stress test at -.165 without issue. I guess I have the best silicone ever made or I'm missing something when I undervolt and its not actually doing it. I followed the instructions so pretty sure the stress test is running with the under volt applied.

Please chime in if you also went this low. Thanks.

 

UPDATE, finally found the limit. I was close, got a shutdown at -.180 in XTU. So, rolled back to -.175. about to run AIDA test at that number, if I have success gonna leave it there for a while and set it up in Throttlestop and also slightly undervolt the gpu. Almost there.

 

As with all electronics, your mileage might vary and some CPUs work great with lower voltage, others are not that good and require a lot of voltage for regular clocks. There is no "standard" undervoltage everyone achieves.

It's great if you reach 170, but you might reach instability at some point, I would suggest keeping a higher value, like -150mv, even if it seems stable at -170.

 

Another quick way of testing the stability is to bench using the TS bench in ThrottleStop, it reports error count if any happened during the bench, but you may have to repeat the 1024M 12 threads bench for 3 to 5 times.

 

I found my laptop(not a GS65) was stable with an undervolt when testing but could reboot a couple of times a day, bad news if you do something and leave it on overnight lol just be mindful

 

My undervolt cause BSOD 2 weeks after I thought I got Stable, this was idle.

Sent from my BLA-L09 using Tapatalk

 

My thought exactly, I dialed it back a little.

 

I ran that test but I'm not sure what to "read" it just gave me a number. Could you explain how it reports errors? Thanks

 

I NEED HELP, Again, almost there.. I hope some experts are reading this. Here's where I'm at. I set up throttlestop with an under volt of -.170 after I had a crash in XTU at .-180 and ran a stress test in xtu and AIDA for 10 minutes at -.175 with no problems, went to .-170 for some headroom. Also under volted gpu by -.04 but have not run and benchmarks, not even sure I need to do that at all. I'll test it later

ANYWAY, HERE"S MY PROBLEM: In throttlestop, when I click the "limits" button, under "ring" I have a yellow "EDP Other" warning. I cannot clear it, its always there. If I click Ring, it just flashes and comes back. I googled and read alot but have no idea what to do or if I need to do anything. Anyone have any ideas why I'm seeing this and how to get rid of/fix it?

Also, in my search for answers, someone else had this problem with the same cpu (he got no answer) and he posted screen shots of all his throttle stop settings. I went through them all and noticed one BIG difference. On the FIVR screen when looking at CPU Core, both our ICCMAX settings are 128 A, when I switched to CPU Cache, I noticed mine said 11 A. The screenshots the other guy posted showed 128 A. I also looked in XTU and that showed 128 A even though throttlestop showed 11 A, so I changed the value in throttlestop from 11 to 128. Can anyone confirm their ICCMAX unadjusted numbers under CPU Core and CPU cache??? I switched mine from 11 to 128 and am a little worried. The only other thing I did related to throttlestop was open my Bios and enabled "speed shift" as it was disabled.

Lastly, changing the iccmax number did not solve my EDP Other problem. From my reading it seems this is related to power limit throttling but I can confirm I never saw any of that in XTU over 2 days of running tests. Please help me, I'm stuck and I am sooooo close to being done with this step of my optimization effort. Thanks!

 

@xLima, are you out there, you seem to know your sh*t, see above post^^^

 

Lol. I am here, always vigilant.

ICCMAX can be edited in BIOS. Search the thread for ICCMAX and you will see FALKENTYNE post on what setting to change to 800 to get an ICCMAX of 200A for adequate power delivery.

As far as EDP other you will still get it, but I want to say not as often, at least not on idle.

Also once you are in That setting you can disable TDC (I assume you Uninstalled Dragon Center)

This is advanced BIOS setting btw. (BIOS unlock key)

Sent from my BLA-L09 using Tapatalk

 

This is how it looks like when TS bench detected errors while benching.

The progress % disappears and error count appears.

 

Thanks, I had no errors

 

No, I did not uninstall dragon center. I like it. Are you using throttlestop? Can you confirm the setting I pointed out should be 128 and not 11? Lastly, can you explain what EDP other means, is it effecting my performance? Its only yellow, not red. I don't want to screw around in the bios too much, just want to stop throttling. Anyway, I may just RMA this unit, see next post

 

OK, need some advice, I've followed everyones advice and am still getting throttling under load. I Undervolted, changed multipliers, changed fan settings, ran all tests with full fans. My temps at idle are 33C BUT when I run a stress test 3 cores are always reaching the thermal limit. Its every other one, core 1,3,5 or threads 1&2,5&6,9&10. They all reach 93, the other cores top out around 85. So, it seems I have a cooling problem with the heatsink or paste. I'm guessing and hopefully somebody knows, that the reason for this is bad contact on half the cpu since its every other core.

So here's my question, would you RMA this unit for repair? or do you think I should try a re-paste? I'm not afraid of getting inside, but would rather not and if its the cooling mechanism and not the paste it would be a wasted effort. My factory sticker is still attached so no one else has been inside. Ugh, this is killing me. Thanks for all your help. BTW, I'm in USA, anyone on east coast send one for repair? Where did it have to go and how long did it take? Oh, and do you think they would take it for repair with this problem?

 

Again, its the QC from MSI. The heatsink and the CPU die don't sit flush so some cores run hotter than other, so when those hit thermal limit, they all clock down. One solution is thermal paste, as far as my testing with CM MasterGel Maker Nano, it fixes it to a large extent but sadly the repaste is optimal for 2 months, then it goes back to how it was.

You may have more luck with a different type of paste, you need to use a thicker thermal paste, I have not personally tested Kryonaut but was told it would not be ideal for the setting because it isn't as thick. Maybe Phobya Nano Grease. LM I have not tried due to my fear of creating a short, but man people here have and has performed perfectly and provided lower temperatures.

As far as RMA it is really your call.

In regards to ThrottleStop and Dragon Center, it is recommended that you don't run them together because Dragon Center does its own tweak to the BIOS and may adjust or modify ThrottleStop changes, also Dragon Center changes TDC value in BIOS.

You can uninstall Dragon Center and install Silent Options for fan control. I personally do not run high performance mode because that forces the CPU to run at max speed always, I rather it max out when it needs to. I use a very customised balanced mode setting. The last to things mentioned can all be found here on this thread, as well as the information backing these claims. (Try searching TDC and 'Sheep', as well as ICCMAX)

 

Hello @ mickey j,

Can I ask you if the offset you apply via Throttlestop on your core and cache remain even after a reboot or shutdown, without running TS again?

I noticed on my GS65 the are still applied, with TS not starting anymore.
The offset revert to default 0 value only by forcing restart or shutdown with SHIFT key pressed.

Me too I've keep Dragon Center installed, but to be honest I have no reason to keep it; i dont found any utility now with it.

 

@mickey j

https://imgur.com/a/Ff06gag

 

Hello @xLima!

Can I ask you why are you using "NVIDIA Inspector v1.9.8.7" instead of "NVIDIA Profile Inspector v2.1.3.19"?
I am really confused about the difference!

 

Oh haha I didn't know there was another. Just Google and downloaded the first result.

Anyway it does what I wanted, and programmed it with task scheduler. Was the only work around to the Optimus glitch with MSI Afterburner other than programming task scheduler to auto start and auto close MSI Afterburner after setting OC.

Sent from my BLA-L09 using Tapatalk

 

Thanks for this. Sorry to be a bother, but again, when In throttlestop, does your cpu core and cpu cache iccmax number match? I see you booted yours to 200A in CPU Core screen, is it also 200A on the CPU cache screen? BTW, I'm gonna try the repaste. I'd rather do that than leave it up to someone else. I'll report back once complete. If that solves my dilemmas I'll be done. If not I'll uninstall dragon center, etc. The reason I want to keep dragon center is because I can easily overclock my gpu rather than doing it the hard way.

 

I have yet to go through the process of creating the task to make throttlestop run at startup which is required for the offset to be applied (about to do this). here's the link: http://forum.notebookreview.com/threads/the-throttlestop-guide.531329/page-454#post-10247294
Once I manually start it, the offset is applied. Hope this helps

 

Thank you mickey j.
On my GS65 the offset keep alive after reboot even if TS is not started (no task or manual start).
That's why I asked.

 

They all match, if you look at the top right white window in Throttlestop you can see the offsets for all the options.

ICCMAX does not appear.

I modified my ICCMax directly from BIOS.

Sent from my BLA-L09 using Tapatalk

 

I'm in doubt between the GM501 and the GS65, what do you say? Has anyone had the 2 and decided by which?

Another question would be in the choice of VGA, whether 1060 or 1070

 

You should wait until the RTX laptops come out so you can get the latest stuff. If you’re still adamant about getting a 1000 series GPU, then you’ll also get a discounted laptop.


Sent from my iPad using Tapatalk

 

Unfortunately I can not wait, because I am from Brazil and will be in the US until Jan / 19. And importing is out of the question because of the high tax rates. So I'd rather take with what you have today.

 

With RTX have to wait and see some thermal testings first. I noticed from CES a lot of company left same chassis but only change GPU.

I would recommend gs65 if you are not only gammer for battery life etc.
GM501 is not bad option in my opinion for pure gaming.

Based on reviews gm501 stock thermals are better then gs65

I am owner of gs65 1070 maxq and I love it no issues at all and all AAA games run perfect.

 

New model with RTX280 MaxQ listed for $2799 US:
http://www.hidevolution.com/custom-...5-15-6-fhd-144hz-i7-8750h-rtx-2080-max-q.html

RTX2070 MaxQ model for $2399 US:
http://www.hidevolution.com/custom-...4-15-6-fhd-144hz-i7-8750h-rtx-2070-max-q.html

RTX2060 model for $2099 US:
http://www.hidevolution.com/custom-...lth-007-15-6-fhd-144hz-i7-8750h-rtx-2060.html

Expected release is set for Nvidia's CES RTX Laptop announced Jan 29 date.

It seems odd to me, but it shows GS75 models as exact same price for sending specs. I would have thought the larger screen would have resulted in a higher price (though I prefer the higher portability if the GS65 myself).

Peter

 

They’re all estimated prices for now. The real price should be revealed closer or at launch


Sent from my iPad using Tapatalk

 

The only 2 companies that showed any improvement in cooling is Acer and MSI.
MSI has 6 heatpipes now for the GS65 instead of 4 from the 2018 model. The GS75 has 7 heatpipes. Both models have improved fans as well for more pressure and CFM.

The Acer is using a new generation (4th gen) fan design to improve cooling.


I would recommend the GS65 as well. For the GTX 1060 or GTX 1070 Max-Q, it depends on your budget. Personally, if you plan on keeping this laptop for a while, get the 1070 Max-Q. I had a GS43 with a 1060 and the GTX 1070 Max-Q gets more FPS, especially when it’s overclocked to non-max-q levels.

 

Does anyone know if they have improved the rigidity of the case? I'm torn between getting the new rtx 2060 Razer and this one .

 

Unsure, but I’ll let you know once mine arrives lol


Sent from my iPad using Tapatalk

 

For anyone else struggling with this major annoyance, the workaround is to use the keyboard remapping functionality build into the SteelSeries engine app. I didn't know such a thing even existed, but it allowed me to remap my right CTRL key to the Pause/Break key (under the "COMMANDS" section/group), so now I can hit LEFT_CTRL+RIGHT_CTRL to issue a "CTRL+BREAK" command. Hooray!

 

Definitely do, I'm very interested as MSI has always treated me well. I like the asthetics of the Razer but i feel their driver support is lacking.

 

I bought the gs65 1070-maxQ by amazon. And now, what are the best known problems and improvements I can make in the notebook?

 

I suppose you need to check your temperatures and confirm if you need to repaste. Users also tend to work on undervolting to keep CPU temperatures in check and avoid power and thermal limits as much as possible.

Enjoy your new laptop

 

I ordered the 1tb ssd. Does anyone know this ssd? It's got a great price

https://www.ebay.com/itm/HP-EX920-M...374689?hash=item590c0ea121:g:EUgAAOSwo~la7a06

 

I;ve read some reviews about the line and its a fine SSD. I think several brands already have great products, its just that some are really great performers.

Hell, even regular SATA 2.5 SSDs are really good lately. It should be fine for more cases.

 

For those who have followed and replied to my posts, I'm talkin to you @xLima , you already know I've been on an under volt mission and reached the limits with temps still too high. I debated sending out for repair but decided to do a re-paste myself. I am now done. I was meticulous and extremely careful opening that thing up. I'm actually on my old laptop right now, I'm afraid to plug in and start the MSI (i will shortly). Anyway, cross your fingers!!! I hope I solved the problem (which was highly variable temps on different cores under load, like 15C difference). So, while I had it open I took some shots of the paste job in case anyone was curious, I'll update you once I work up the nerve to plug in the stealth.

 

^^^As you can see, the paste job on the cpu was crap! It was also hardening while the gpu was still soft

 

Did the paste reach the thermal pads? I see that they are grayish.

Also do not get scared computer will not boot if you do not plug it in.

Sent from my BLA-L09 using Tapatalk

 

Yes it did, I left them alone.

 

I am now typing on my Stealth, so Success as far as not killing it. After bootup and regular fans running (30% and silent) my temps are 35C. Time for a stress test. Gonna start with the XTU test for 10 minutes, full fans.

 

This is with ARTIC4?

Sent from my BLA-L09 using Tapatalk

 

OK, ran my stress tests and benchmarks with mixed results. BTW, I used Artic Silver 5 for pasting, why? because its what I already had.

XTU- no throttling running full fans. Top temp 80
AIDA-throttled, but less than before. Still experiencing every other core running way hotter than the others, ie, 1,3,5 (scratching my head on this). However "cpu throttling max" was 26% after undervolt but before paste. Its now 9%
-^^^both of these ran 10 minutes
-My cpu under volt is -.170, so far no problems and I just ran a bunch of tests back to back to back.
-My GPU under volt is -.04, again no problems. I just guessed on this number.
Throttlestop TS bench- no errors on all 3 tests and no throttling but came close, can't remember. Also, for the first time, I just opened "limits" and have no warnings or errors.

Using 3dMark to test GPU running the timespy benchmark keeping in mind I have the 1060
-I believe I ran my first test after some under volting but before the paste. My score: 3813
-I ran several more tests, my best overall score included running in performance mode in dragon center, with full fans on and overclocked at core gpu at 65 mHz and Vram at 130mHz which is about 1/3 of the limits. Score: 4018. Note that even though this was my best overall score, I got a higher cpu score at a lower overclock.
No throttling or graphics problems although I think I was getting close to screen problems. I did throttle when fans were not on full but very briefly.

I have screenshots of a lot of these results if anyone is interested.
I don't think my results are as good as others but good enough for me, just wish I knew why 3 cores run so much hotter than the other 3. I hope my cpu isn't bad.

Please chime in on what you think of my results and any advice you have, etc. I'm here to learn

I'm done for now with this type of stuff. Time to go through services and apps and settings and kill all the crap or unnecessary stuff. I've done a little, much more to go. Also need to install some programs.

Last thought is a question. Does anybody know if the MSI backup program will save all my settings as they are right now or does it save a factory version? I'll run a full image backup anyway but was curious. Peace!

 

The problem with Artic Silver 5, is that it takes several hours for the paste to reach its optimal state, you have a significant cure time unlike pastes like gelid Extreme or Kryonaut. And in my own experience, I have more trouble correctly applying AS5 compared to Kryonaut where I just have to put a pea in the middle.

 

It saves to a factory reset state. So no, your current setup will not be saved. You can try macrium reflect to make a backup.

Post Firestrike score.

Sent from my BLA-L09 using Tapatalk

 

It's been exactly a week since my CM MasterGel Maker Nano repaste and I've already seen my maximum temps go up by 2-3 C in the same room, same ambient temperature, with the same fan settings, running the same tests. I'm not quite sure if this is part of the "settling in" period of the paste but I might have to look for a different option at this rate.