*** The Official 2019 MSI GS65 Stealth with RTX GPUs Owners and Discussions Lounge ***

Go to the Nvidia website and download the DCH driver.

 

Whats the difference with DHC?

 

It's the new Driver Standard from Microsoft.

 

I'm curious which laptops have a 90w 2080 Max-Q.
I'd like to try and get that bios and potentially flash it on my laptop...

When I used to build desktops, I would occasionally flash a new bios to the GPU to increase performance (R9 290X).

 

Someone versed in tweaking GPU bios should be able to make a modified vbios for these machines. I will pursue this with the GS75 eventually, as it seems to have enough headroom. I might just try capping it at 100w to see what it can do

 

Right now, Fire strike shows 13276 (graphics=14689, physics=15069, combined=6989). I just have the stock version gs65 2060. I will run the Time Spy again to find the individual scores, and will play a game for half an hour later today...

 

yes, the latest nvidia driver from the website for the 2060 notebook does not install on my gs65 2060. I have to use the msi provided driver on their website.

 

some more gaming benchmarks:

assassin's creed origins:
gs65 2060: 5500 benchmark score
for comparison...
15r4 1070: 7000 benchmark score

rise of the tomb raider:
gs65 2060: 61 fps benchmark score
15r4 1070: 83 fps benchmark score

 

I wonder if something is limiting the RTX2060 there. I would have imagined it would perform closer to the 1070 all around.

 

Can you please do the Port Royal benchmark for RT on your 2060 laptop GPU? I think there is also a new DLSS "test" built in to that benchmark, but IDK if it's got it's own DLSS benchmark result, but if it does can you please run that and post it to? Thank you!

 

I wish. I love the form factor of the gs65.

Both systems have fresh installs of windows (I reinstalled windows on a partition of the gs65), and the very minimal drivers needed to run those tests (latest nvidia drivers, etc). Both have identical hardware specs other than the graphics cards.

On the fresh install of windows, I can get 25 more time spy points on the gs65 2060. It comes out to:

time spy=5497 (graphics=5555, cpu=5192)

The time spy benchmarks with 3dmark are somewhat close to the 1070, but when I run games, the performance is about 70%-80% of the 1070.

 

unfortunately I cannot because msi has not updated their own nvidia drivers. those are the only ones that will install on my system. if I try the latest ones on the nvidia website with DLSS, it says my system is not compatible. I am guessing the latest driver can probably be installed on other 2060 systems.

 

Interesting, I would have thought MSI would have gotten those ID's to Nvidia to include in their driver updates by now, probably the next one, maybe. Please give it a shot when you can, thank you!

The Port Royale RT benchmark should already be working, even if the newest DLSS test isn't working due to new driver being needed?

 

I am checking the Port Royale benchmark, and it says I need to pay for it, the advanced edition of 3d mark.

 

It was on sale yesterday, from another post:

 

Hi, yes...I do not mean to be rude, but I can do all the free benchmarks no problem. That is all I really feel comfortable with, if you would like me to do any other tests.

 

The point of interest is that Nvidia has reduced the Gigaray's throughput of the RTX laptop GPU's in relation to the RTX desktop GPU's, and so it's of interest to see how much of a hit the 2060 laptop RTX GPU has compared to the desktop, here's the chart of Gigaray's throughput for desktop / laptop RTX GPU's:

The Port Royal desktop RTX results, showing an RTX 2080 laptop scoring closest to a desktop RTX 2060, so given the 2060 Gigaray's have dropped from 5 to 3.5 (High and Low), the RTX 2060 laptop GPU will be hit hardest.

Given the poor RTX performance from desktop GPU's, it seems clear the laptop RTX GPU's aren't going to be worth much for RT RT, and it will also be interesting to see how the reduced performance affects DLSS.

RTX arrives in laptops
http://forum.notebookreview.com/threads/nvidia-thread.806608/page-173#post-10856786

PCWorld found RTX Off = 90 FPS, RTX On = 45 FPS, on newest drivers / game update on a laptop RTX 2080 MQ - GS75

Hopefully someone will run the 2060 / 2070 laptop RTX GPU's against Port Royal so we can fill in the graph. Thank you!

 

If you go for this, keep me in the loop. I'd like to flash my bios to 90 watts =)
Do you know how to mod a vbios? Or have someone that can?

 

Sure I will keep you posted, and I don't have anyone or someone that I know of, yet. That has never stopped me mwuahaha.

I am sure there are people in this forum with the capability to do so, even if temporarily. We will find more in time, don't worry

 

So I tried that no luck, will just try DDU and reinstall GeForce experience

 

Ah I'm getting this same issue as well glad someone else has noticed this

 

I think I only mentioned this in another thread, but every new gen and system, there is always some hiccups for a month or two with GPU drivers. I have experienced these issues every single new machine since 2012 haha. Don't sweat it

 

Thanks @ryzeki thats a relief to know!

Now does anyone else not use dragon center? I was looking into using silent option but it hasn't been updated for some time it seems and doesn't look too different from the one included.

 

It works with these laptops.


Sent from my iPhone using Tapatalk

 

Yeah I saw your review, great stuff! So with Silent Option any particular reason over dragon center? I saw before you'd mention throttlestop issues but are the functionally identical?

 

If all you want from dragon center is the fan control, then silent option is well... an option

Thanks for confirming it works. I do plan on using it and I would rather just use that instead of full dragon center

 

Dragon center also incorporates CPU controls and those overlap/fight with other software like Intel XTU and throttlestop so you shouldnt use both simultaneously. Dragon center itself is the sum of independent software, so you can use independent software to achieve the "same"

You don't need Dragon Center itself, it just neatly incorporates all the rest into a single UI of sorts. I personally would use MSI True Colors and Silent option independently.

 

The biggest feature I like of Dragon Center is the battery restriction settings, it's convenient for the longevity of the battery. ex. say if you're gaming plugged in you can cap the battery at 60. Is there a software that does this stand alone?

 

You mean changing from "performance" to "eco" or something like that? (Sorry it's been a while since I used Dragon Center). If it is, I believe this was built into the SCM (system control manager) that handles the FN+F# keys) so in theory you can switch on the fly those without Dragon Center.

Unelss new machines no longer use SCM?

 

So out of curiosity did yours come with dual channel RAM?

 

I don't have SCM downloaded but I could check on that because if so I'd happily remove dragon center

 

It would be great if you could check for me, because at least for my machine, SCM is a must, as dragon center requires it to run.

 

It did not. The Alienware I am comparing it too also only has 1 stick 16GB ram, so on my end, I was just concerned with even comparisons since I am looking to upgrade.

 

I am pretty stoked about the new MSI GS65 Stealth-005 and just ordered one from HIDevolution (Donald was amazing to work with).

In looking at notebookcheck, I see that in the comparison the screen is significantly brighter in the SB2, however, the brightness (to some extent) has to overcome the glossy screen.

Has anyone had any experience coming from the SB2 to the GS65 that could shed some light (pun intended) on how the screens actually compare between the two side by side?

 

I just got the GS65 with 2060, 16GB Ram (Single Chip) and 512GB SSD. Out of Box, some Windows updates and installed 3DMark along with HWINFO. Besides TimeSpy not knowing what type of Video Card I have, scored a 5490 Overall.

https://www.3dmark.com/spy/6108254

Only bad thing is my termals are hitting high 80's with Core #2 maxing out at 91. Dragon Center set to Performace profile.

Compared to what some others have posted regarding Thermals, seems this is pretty high.

Will try Undervolting with XTU to see if I can bring the temps down.

 

What’s your ambient temps?
Where did you buy the unit?


Sent from my iPhone using Tapatalk

 

My office is currently 71F and it was purchased from Best Buy. I also have 2070 with 32GB ram coming tomorrow from Amazon.

 

Welcome to the forums! The GS65 has gathered a lot of love in general, and seems to be a great product all around, particularly with the improved thermals and performance of this current 2019 model. Hope you can give us your impressions once your machine arrives.

Undervolting will definitely help. Your max temps will decrease and your sustained performance will increase. Factory paste is kind of like a roulette so sometimes you can get hotter than average temps. however.....

This is very important, and many users tend to forget we all have different ambient temps, and they are very influential in some cases.

 

I have another question. I also purchased a Razer Blade 15 2070 and notice quite a difference in CPU scores on the TimeSpy benchmarks.

The Razer has a CPU Score of 6664
https://www.3dmark.com/spy/6086744

and the GS65 has a score 5220.
https://www.3dmark.com/spy/6108254

Both are stock and not undervolting.

Not sure if it's MSI doing something to limit the CPU or if it's something Razer is doing.

Anyone have any thoughts?

 

interesting as I'm unsure why my score was so low... I'll re run it again tonight and I'll also purchase 3D mark as well while it's on sale

 

This is good information, now we have one clue as to what could be happening. Honestly we don't know yet. Can you run benchmarks with monitoring software to check CPU usage, speeds and all? Maybe under stress, something in the EC was throttling the CPU and thus ended causing lower performance there.

 

So according to the MSI updater UWP app, and a quick windows search I do not have SCM installed but have Dragon Center 2.0 installed

 

Undervolted both laptops -0.125 in XTU and ran side by side.

Razer Scored 1469 Marks, Max Freq. 3.95Ghz and Highest Temp 78C

GS65 Scored 985 Marks, Max Freq. 3.91Ghz and Highest Temp 91C

TBH, not sure what else to look at in HWInfo. I did notice that the Razer was Throttling Power Limit from Time to Time.

 

Tomorrow when the 2070 comes in I'll run some more benchmarks. I'd really prefer to keep one of the MSI's.

 

Interesting, maybe SCM is no longer independent and comes with Dragon Center... someone else can help here, because I can't until I get my own machine haha.

How odd, we would need to graph clocks, usage and any possible throttling during the duration of the tests. I suspect the MSI was throttling, and might need a repaste looking at those temps.

 

Yeah, I want to see how the 2070 compares to the Razer before I go down the repasting road. I have 32GB RAM and (2) 2TB 970Evo's to install so I'd do it all at the same time.

 

Oh man, I wish I had 2TB Evos to install haha.

 

Remember that some of the more active posters here may be running LM and non-stock thermal pads.

Sent from my BLA-L09 using Tapatalk

 

Hmm I will need to double check my scores as well because It seems that my result was a bit of an outlier. I'm going to run it again when I get home.

 

This is generally false on Covfefe lake mobile with respect to temperatures. 45W is not enough to sustain the turbo except in rather low CPU loads, you need 60-80W on average load-dependent. This means that undervolting only lets the power limit stretch further. You might go from 3.2GHz to 3.6GHz, but you will still be hitting 45W constantly under any meaningful load like a CPU heavy game such as Monster Hunter World or CoD: Black Ops 4 (which uses AVX workloads I might add) which will produce the same temperatures. It's just like overclocking Pascal mobile; you're still hitting a power limit in most games so your performance increase is essentially "free", but you don't drop temperatures.

It'll help if your GPU is too weak to sustain high FPS in a game that's already fairly CPU frugal, like... Fortnite, but in that case you'd likely already be in safe temperatures on the CPU anyway, and undervolting wouldn't actually improve performance since the CPU is clocked up fast enough to begin with. It's one or the other. Either you drop temperatures and keep exact performance, or you gain some performance and temperatures do not change. Sadly very unlike the older CPU generations.