XPS 15 9570 Owners Thread

Sleep isn't just about how fast it boots. For work I typcally have many windows open and a certain workspace, takes quite a while to restore everything if I shutdown. So it is quite important to me.

Yes reliability matters. With the 9550 it is almost reliable after I turned of that "Hibernate after" that made it wake up in the backpack instead (nice feature generally, but unfortunately not reliable).

I still get an unkillable ram-eating runaway windows process if I try to connect to my e-banking site using a certificate after the laptop has been sleeping!?

 

Here is a good example btw, how GREAT modern standby is (works... not works):

https://forums.lenovo.com/t5/ThinkP...n-6th-Gen-Battery-drain-in-sleep/td-p/4075415

17 sites full of hate and it is basically the same issue as I see and have it. Lenovo doesn't seem to care too same as Dell. It's not our problem right, Microsoft wants modern standby, so we had to implement it. No one ever said though, the laptop manufacturers had to remove S3 totally. It just makes no sense, to implement modern standby on a laptop.

I had not a single device so far, where modern standby worked reliable over a long time period. I just think the Microsoft own devices like Surface have a "proper" working modern standby, though there are hundreds of complaints about it too, and a very long thread on the MS forum since now over 2 years, with a hibernate issue.

The problem with modern standby is, that it is no "real sleep" modus. The laptops is basically still on, the OS just tries to put it into the lowest sleep state as possible. Every little software running on the OS can, and will, work against this though. Driver issues then come together with it too of course. Because the computations never really stop, the whole thing cant work. Thats also why if you have running, for example, a game while you put it into modern standby, it wont really "cool" down and will always run with fans on and also stay warm or even hot and drain energy a lot of course.

A laptop is a working instrument, you should rely on, at work, university, ect. And you want to be mobile with it. You maybe have a lot of work running and open on it, maybe even CPU intensive tasks. And then you want to put it to sleep, and go somewhere else with it, and during that sleep time... you want it to be OFF.

And S3 works perfectly for that, it worked for years. Modern standby is just a horrible design, Microsoft is too stubborn to acknowledge, it's a failure since it was introduced. Even the wake up time from S3 is btw just 2-3 seconds... compared to 1 second of modern standby.

The only """advantage""" MS holds for modern standby, is the stupid connected standby thing, so it could theoretically update emails in UWP apps while it "sleeps"... however.... the 9570 doesnt support connected standby

 

This is what i do when people put electronics that are still on into an enclosed space. Just not very smart.

Understandable. I do most of my heavy lifting on my desktop anyway which stays on 24/7 and just remote into it.

 

I was experimenting a bit with the modern standby in the last few days. It actually works great for me!
I was trying to wake it up multiple times with the Logitech mouse and its USB receiver. Looking at the powercfg /sleepstudy it shows the laptop wakes up when the mouse is clicked, but it always goes back to sleep after 30-40 sec because of "video idle timeout". It consumes peanuts during that period, only around 50 mW. Which is fantastic - my old 4th Gen Lenovo Yoga laptop with S3 sleep would just wake up and stay awake, no matter the cause of wake.

It looks like the connected standby puts the laptop back to sleep very quickly (in case it wakes up) if the lid is closed. Which is what we want. That's great.
Bluetooth mouse activity doesn't wake it up, only the USB receiver.

I don't play games and don't have any, so can't start a game and try to put the laptop to sleep to see if it continues to run, but I'd be willing to give it a try for any other scenario.

So far, what I've tried, the modern standby works much better than the old S3 sleep FOR ME. I'm comparing it to several older laptops I've had in past. The old S3 sleep was never very reliable for me, even when I fiddled with them, trying to find the causes of wake ups. Any damn Windows service or any program or trigger can wake it up, and when it awakes it stays awake...

So, if the new modern standby puts the laptop back to sleep if it wakes up with the lid closed then I'll gladly pay the price of 1-2% of battery drain (that's over the period of several hours), over the old S3 sleep.

 

This forum...

@ splus please stap :/ the mW value in the line is wrong, you have to click at the end in the last line. It will mostly if everything works "correctly" drain terribly bad 600-1000mWh, and if it's not 100 100, then drain up to 2500 or even more. And what are you even talking about ... nothing you say makes any sense. Nothing. Not a single sentence is correct. And S3 is just as "reliable" as how smart the user in front of the laptop, and how well enough he configured Windows. Per default Windows will wake up from S3, you have to deactivate that. It is not too hard to learn to use Windows properly.

Microsoft allows per definition btw just about 0,35% (it's in the tech docs somewhere for modern standby) drain per hour of total battery for a CONNECTED standby device.

 

Sleep and wake up has been wonky on at least 3 generations of XPS laptops.

I gave up on it and don't use sleep.

 

I've just received my 9570. I'm quite concerned about how dim the top quarter of the screen is compared to the rest of the display. I know some uniformity differences are to be expected but this is very noticeable on all screens and distracting.

Does anyone else have this?

 

The battery consumption in sleepstudy matches the total battery used during the period of testing. I've checked the battery level with 2 decimals before going to sleep and after.

What do you mean "how smart the user is" for the S3 sleep? A new app that doesn't play well with the sleep of a particular hardware or even a Windows update installs some hooks or triggers and it can upset BOTH the S3 sleep and the modern standby.
Should a "smart user" test his laptop every few weeks and comb through all possible Windows events or any possible settings to see if and where a possible cause could be? NO user should do that.
S3 sleep has always been VERY dependent on a particular hardware. Some laptops sleep pretty well, some don't. And looking at that ThinkPad X1 thread shows the same problem is with the modern standby.
S3 sleep is nice in theory. It has never been so nice for me in practice. Most of the time was OK, but let's say 1 in 20 it wasn't, the laptop would wake up for whatever reason, some app or whatever else...

 

Yes, this forum. Quite a few condescending people who think they’re smarter than they actually are.

 

Bunch of fart smellers

 

Has anyone tried to disable the touchscreen in BIOS to reduce the power consumption?

Most people say that in theory it shouldn't matter much, but people who have tried say they get noticeably better battery life.
I have a 4K panel and don't use the touch at all, so I tried to turn off the touchscreen in BIOS. I can't be 100% sure but think the idle power usage seems to be gone down a bit.

On another note - my keyboard was missing keystrokes, around 1 in 20, which was reeeeally annoying. I thought it's a hardware issue, and was already thinking to call Dell to replace the keyboard. But then I looked online, and it looks like it's a common software problem with quite a few Dell laptops, especially XPS and Precision models.

The BIOS fix for missing keystrokes on 9570 is apparently in the works, but someone posted a temporary fix - disable the Intel Panel self refresh option in Intel Graphics power settings. Sounds weird, but it works!

Panel self refresh option is a good power saving feature, which causes the display to refresh less often if nothing changes on the screen. In my previous Lenovo Yoga laptop this option would reduce the power consumption by almost 1W. However, I couldn't see any difference in XPS 9570 with this option on and off.

 

A question:
Has anyone with a colorimeter made a calibrated color corrected profile for XPS 9570?
The reds are quite off on a 4K display...
Edit: Actually, both reds and blues seem quite dark, whereas the greens are much lighter.

 

There seems to be quite a lot of variation between the panels despite the same label. I once tried a profile posted with a Notebookcheck review, but it wasn't good. If you don't want to buy a colorimeter you can try borrowing one or having the laptop calibrated somewhere. Mind older devices are not good for wide-gamut panels.

 

I have a Dell XPS 15 9570. I want to change my PC401 NVMe SK hynix 1TB M.2 SSD with a Samsung 970 EVO 2TB M.2 PCIe drive. The PC401 is set up as a Raid Drive on the Dell. I want to clone or copy over to the Samsung 970 everything on my Dell PC401 Drive. I’ve tried Acronis True Image but the Dell doesn’t recognize the Samsung 970. Is there a way I’m supposed to set up the Samsung 970 drive first? I’ve also tried Snapshot and Macrium Reflect. How do I get the Dell XPS 15 to see the new Samsung 970 EVO drive?

 

Hi everyone, been lurking on this thread for a bit. Received my 9570 a few days ago, basic specs: i9, 32GB, 2TB.

I'm a keen amateur photographer and my main computer is a custom desktop (Haswell-E 5930k). I use Photoshop, Lightroom, Premiere Pro, etc. For my laptop, the objective has always been something that combines power with portability. Ideally, it would be something that comes close to the performance of my desktop.

Until recently my mobile workhorse had been a 2012 15'' Macbook Pro i7 3820QM. However, I've been feeling the Mac no longer cuts it. It's old and feels a bit slow. I'm also bit fed up with Mac OS these days. It seems clumsy and dated, so decided to go Windows. Hence the Dell.

I'm aware of all that's been written about throttling and have played with under-volting and so on.

But how does it do in the real world? With that question in mind, I set up a 4-way test of my desktop, the Dell, the Macbook it just replaced and my other machine - a Clevo gaming laptop (Skylake i7 6700 desktop CPU).

The test was a 150 photo Lightroom export. Most of them are in RAW format (needs hefty processing and memory) and all have various effects applied to them.

The desktop was heavily overclocked to 4.3GHz (all 6 cores) and has liquid cooling. Some would say that's not a fair test as none of the other machines are overclocked. I'd disagree. The thermal advantages of a desktop with proper cooling are inherent to the format. The other machines can't be overlocked even if the CPUs were unlocked, because there's no thermal headroom. (In any case I ran the desktop benchmark twice - once overclocked the other time at stock settings).

Here's a rough and ready video of the test. Hope you find it interesting.

I think the Dell turns in an impressive showing, considering it's up against a reasonably powerful 140W TDP CPU from not that long ago. Stock vs stock, the results are as close as makes no difference.

 

My guess is that TI needs to see the NVME driver in order to recognize the new drive to accomplish the cloning process. Go to the Acronis Forum for True Image 2018 and search for the MVP Tool-Custom ATI WINPE Builder and follow those directions to build a bootable TI usb/dvd that has all the drivers you will need in your XPS. Unfortunately the standalone Acronis Software while good doesn't have all you need to boot into a modern computer and do its thing well.

 

Have the same version of 9570/ssd, looking forward to see how much speed you gain with the 970 evo. I have been looking for an external NVMe SSD Box, but they are hard to find and very expensive

 

Seriously, what is up with the boot up problems? I though perhaps my machine was again the victim of poor QC from Dell but I get the impression it is a common problem for the 9570 to hang during boot up/post?

 

Are you refering to seeing it in the BIOS or in Windows?

 

I've a calibrated profile for the FHD monitor!
Today Dell SupportAssist try to update the integrated graphic card to 23.20.16.5037 driver version, but display error message and ask me to get driver from manufacturer website. Yesterday I've noticed red dots in the black area of pictures, using Google Chrome with hardware acceleration.
Is anyone else experiencing the same issues?

 

What is the best thickness of thermal pad to get for the interior of the 9570? 0.5, 1.0 or 2.0?

 

I dont have 'Force S3 Sleep' in the bios. .where you find it? I kinda remember i saw it before, but i no longer can find it..

 

You can't anymore. Dell removed it with bios 1.3.0 because they're a bunch of losers and scammer. http://forum.notebookreview.com/threads/xps-15-9570-owners-thread.817008/page-81#post-10773467

You can still just deactivate modern standby per Windows registry, but I dont recommend it, because the CPU is bugged after S3 resume and consumes way more power than it should do, and Dell dont want to fix it.

This laptop is just a total joke.

@Paul Dubey you have to switch to AHCI for a Samsung SSD to work... and use the Samsung nvme driver.

 

I did some testing on the sleep mode. Here are my results:

1) Modern Standby Power Consumption: around 650mv/hour (firmware 1.2.2)
2) S3 Standby Power Consumption: around 400mv/hour (firmware 1.2.2)
3) Entering modern standby while the system is under stress. I had no problem here.
I was running heaven benchmark for about 10 minutes and then entered standby without closing any program.
The screen went blank immediately. The fans kept spinning for 5 seconds at full blast, and another 5 second at reduced speed, and then complete silence.
Upon resuming, I verified that the system entered standby and the power consumption was the usual 650.

@maffle Have you noticed that your screenshot reports a rate 2mW/hour which is much less than my 650mW/hour (the battery is 97000 mW).
Maybe you have a problem with how windows is reporting things? I don't know.

 

I'm not sure if this has been answered yet, but I'm on the fence between keeping my XPS 9570 with the i9 or switching the an MSI GS65-259. Right now, I've LM'ed both dies and temps overall are pretty good with cpu hovering around 80 running 4ghz. The issue is that after a longer sustained load my clock speeds drop to 3.4ghz and throttlestop says that I'm being limited by PL1. This is kinda of annoying because, while its still fast, I borrowed a GS65 and it throttles down to 3.7ghz under the same load, which means that I'm literally running slower than the i7. I'm not really sure what PL1 is, and I can't think of any way to remove it as I've already slid all the power settings of intel XTU to the max.

 

The SATA setting in the BIOS really perplexes me.
I did some digging into the whole AHCI mode / RAID mode situation.
Let's start with this: the SSD drive in the 9570 is modern NVME/M.2/PCIex4 drive. It should have nothing to do with SATA/AHCI. (Yes, there have been hybrid solution like SATA express but this is not the case). This should be a native PCIex4 with no contact with the SATA subsystem and the AHCI interface.
So why does the BIOS even bother asking us which SATA mode we want?

I am inclined to believe that
1) RAID ON - This enables an Intel "Chipset-level" Raid implementation that handles *BOTH* SATA/AHCI disks *AND* NVME disks. SATA disks are handled with a generic intel AHCI driver; NVMe disks are handled with a generic Intel NVMe driver. You can do this chipset "soft" raid if multiple disks are available.
2) AHCI mode. Does nothing in our case because we do not have any SATA/AHCI disk!
What happens when you select 2) is that the SSD is (correctly) using the NVME interface (not SATA/AHCI) either using a generic NVMe window driver or a specific driver from the Manufacturer if available.

This seems to be confirmed by the name of the drivers/controllers in windows' device manager.

To summarize, when you select AHCI VS RAID on, you are actually selecting between NVMe drive handled by an NVMe windows/manufacturer driver, or NVMe drive handled by intel RAID controller and intel NVMe driver.

If anyone has a better understanding please do share.

 

@ApplesOfEpicness Did you pad the ram sensor?
Check out this thread http://forum.notebookreview.com/threads/dell-xps-15-9570-benchmarks-temps.817970/
It would be nice if you posted some of your benchmarks (like realbench) since you are the first I see with an i9 + LM.

 

PL1 is the power limit throttling quite typical for XPS 15, implemented via Intel DPTF. Presumably on the 9570 it tends to kick in when the DIMM sensor of Dell EC is at 63 (using HWinfo64), could also be due to other sensors/thresholds. The root cause of this is likely heat from the VRM mosfets which have no active cooling by default (the first Ambient sensor is nearby, what does it read)?

To work around the VRM heat, folks use a couple of different ways of padding the mosfets, which are simple to apply but of limited effectiveness. You seem to be after ultimate performance, so you should look into iunlock's mod for the 9560, which essentially adds internal cooling surface for the VRMs, and redirects some air over there
http://forum.notebookreview.com/thr...ures-benchmarks-xps-15-9560-kaby-lake.802345/
Not sure if this exact approach is actually needed, you might try something easier and non-destructive first. Mind iunlock tried quite a few simpler ways before he came up with this one. But he had to get the first Ambient to lower seventies for the 9560, and I think GPU undervolting via the voltage curve in MSI Afterburner wasn't practiced yet then.

The other thing is padding ram and the nearby DIMM sensor to the backplate. This should have a more direct effect on reducing throttling (assuming you don't heat up the backplate with VRM heat too much). But make sure you do at least something to reduce the temperatures in the VRM area first, because they appear to be excessive and some report freezing/shutdowns under heavy loads.

Some have fought PL1 throttling by killing Intel DPTF. I think this may not be the best idea, because even if you do take care of the VRM heat firstly, something may fail at one point and you're running with this thermal fuse shorted. While padding the DIMM sensor to the backplate only offsets the measurement by a few degrees (considering that the threshold is not adjustable).

 

Please remember that there are configs of the XPS15 that has a hard disc, so the BIOS has to support both. If I run the Intel Rapid Storage on my high end config, it seems to indicate that I still have an unused SATA port in the machine. That said, I share your confuision about this setting, as it is very unclear what it actually does.

 

To my astonishment, I noticed that it still seems like you have to disable the Realtech startup programs in order for the XPS15 to detect when a headphone is plugged in, can this really be? Or does it have to do with Wave being disabled also in startup (based on past experiences.....)

 

PL1 throttling at 3.4GHz is not the dimm sensor, that would be more like 1.5GHz... Also having XTU settings for power limits set to the max will cause the CPU to totally ignore those values in most cases. Having said that, the i9 isn't going to be better than the i7 in the XPS.

 

Similar choice here (i'm looking at a gs73 8rf)
I need the laptop for compiling code, using UE4 source as a test. My old 9550 can hold 3.1GHz and 60 degrees, so 20 degrees below stress test levels. I figured the i9 might run cool enough to hold a decent clock.
Out of the box it struggles with power&thermal limits, needed to set Throttlestop at -100/-100 to for 3.4GHz stable. It still hits brief throttles due to core/package hitting 90+ degrees :-(

Not sure what a realistic stable clock would be under this load, i suspect with a re-paste etc it'll still be around 3.4 but running cooler / with less fan noise. Will be interesting to compare to the i7 in the GS73 (i have one on the way)

 

I got some fantastic news about my order


Been 30 days now and still no delivery hahahah awesome !


Asked Dell for compensation. They said "according to our sale contract you agreed upon, we have 30 days to deliver" which they failed.

And the only thing they offer me "please, tell the delivery man you don't want the package, let it go in return. And you can then open up a refund request"



That's the spirit Dell, such an incredible premium customer service!! Mind blowing

 

@_sem_

I don’t think it’s a temp issue. All my ambient sensors look within reason. Most notably, the VRM’s didn’t exceed 80C under cpu intensive loads and the dimm sensors only ever gets up to 60C max. Maybe the limits are lower, but I think that limiting these that much is just a waste.

On a side note, I did run Realbench and came up with some average-ish numbers. The hottest the vrm sensor got was 93. I am looking to improve the cooling for that but am waiting on parts. CPU averaged around 82C and clockspeed bounced around a lot between base and 3.4. I've noticed that I can generally sustain a stable 3.4ghz boost, but thats kind of dissapoint for me because the cheaper i7 boots higher and I do have the thermal headroom to reach 3.7-3.8ghz. The score was definitely limited by the fact that I kept PL1 throttling.

@custom90gt
I know that I probably won't get better perf than the i7, but I'm expecting it to be at least on par because that is literally throwing money away otherwise. (Not that the i9 wasn't already a fairly big waste of money on its own)

 

I know this is an old post, but you can just restart explorer to fix the system tray scaling bug. I added a restart explorer context menu to my right click on desktop context that runs a script to do it. I'm on vacation at the moment for a while but I can send it over when I get back..

In the meantime you can just restart explorer process manually in the task manager..

 

Just wait until you see how long it takes to get a refund from Dell, even though they have the machine back at their premises. I'm sure they do it on purpose to keep your $ or £'s as long as possible.

 

Well PL1 is some sort of throttling (I understand this throttling is not due to the end of the turbo time limit). Not sure how it is triggered, because there simply aren't many reports of folks going into this seriously enough like iunlock did, as XPS is obviously underdesigned. With the i7 9550/9560, running the CPU at max turbo (clock depends on the number of active cores) wasn't such a difficult problem, as long as the NVidia GPU was off. The 6-cores seem to be different animals.

At times, certain throttling thresholds are quite low. It is difficult to say what they are because they don't tell us and there are many numbers on display in HWinfo64. You can try observing what goes on at the moment when PL1 throttling kicks in. Maybe you can save the log with HWinfo64 a couple of times and try to find out with log viewer what temperatures were the same when it kicked in.

Thermal images invariably show a growing heat blob in the VRM area, which is the prime suspect for the cause. Another question is what is the exact trigger.

You can also try adjusting the power limits in Intel XTU up or down and observe the effect compared to stock (you can look up what pressing did).

Maybe try blowing some cold air into the small grille in the central part of the gap between the keyboard and the display part of the laptop when PL1 throttling is active, to see if it goes away.

 

That probably won't work. Dell installed some foam on the "center" of the back grille (the exhaust grille facing the screen) which "insulates the VRM.

So the hot exhaust only vents from fans on the left and right side of the screen.

@GoNz0 explained the purpose of the foam in either the 9550 or 9560 threads but I can't remember. . .

 

I think iunlock removed it. My 9550 grille was half blocked by a piece of conductive tape, which I've adjusted so that it now lets some air from the VRM go out (comes redirected from the fans courtesy of iunlock). If one just unblocks it, cold air can also be blown in for a test.

 

@_sem_

I tried taking a look across a couple of cinebench runs with HWinfo, but I can’t find anything that correlates with my throttling. The only thing that indicates throttling is the, “IA: Package-Level RAPL/PBM PL1” changing from No to Yes. I also tried uninstalling the intel thermal management stuff and it still PL1 throttles which means that I didn’t uninstall everything correctly or that the limit is firmware set which is really pushing me towards the MSI...

EDIT:

Just out of curiosity I moved my setup down into the basement where the ambient temps are like 5C cooler. I also raised the laptop and put a cooling fan blowing on it. Idle temps didnt change but full turbo load temps dropped by around 5C-7C to settle around 78C. However, it still PL1 throttled around the same time after doing more than one consecutive load which leads me to believe that this shouldn't be a thermal issue. As a side note, I did manage to achieve a score of 1279 which is on par with the Xeon X5650 listed in cinebench. Not bad for a mobile processor.

 

I think that redirection of cool fan air and removal of some foam from back of the VRM to exhaust is a smart move. Active cooling should significantly bring down temp of stagnant air in the VRM area.

 

Could it be matter of short/long power limit duration, throttling after a preset time after loading up the CPU? I recall there was such an issue with the new Razer 15 throttling the i7 from 45W to 35W, later addressed with a firmware update. Did you check the power limits with Intel XTU?

https://www.reddup.co/r/razer/comments/8tnrsz
https://linustechtips.com/main/topic/676971-4790k-around-200w-package-power/

> but I can’t find anything that correlates with my throttling.

Likely it would be nothing obvious at first glance, just one sensor reaching the same value in different runs at the point when the flag changes. In case of slow dynamics, the response of this sensor to the throttling may be delayed substantially. Direct thermal throttling acts in seconds, while the PL throttlings we've seen in 9550/9560 rather in minutes.

 

@ApplesOfEpicness
Did you undervolt? If so by how much?
What exactly is your stress load?

 

PL1 is the long time power limit and usually set to the CPU SKU. For i7-8750H@45W TDP. Can't be overridden/increased on locked down processors. If you see lower TDP than Intel's own specs for the processor.... This means only that the Notebook manufacturer has crippled the chips further with firmware / software due lack of creating/design adequate and good enough cooling.

 

Undervolted to -120mv. Load is mixed. Whether it’s cinebench or realbench, it will throttle no matter the load. I think Papusan is right as the throttle consistently hits after a sustained load of over 15 seconds ish. This is all keeping in mind that, while it is throttling, it still maintains boosts around 3.5ghz. That is really good for such a hot chip, but I’m just saying that in my experience the gs65 PL1 throttle seems to bring it down only to 3.7 ghz which is still above the rated 45w top for Intel.

@Papusan you say that I can’t change these limits on a locked processor, but the core i9 is an unlocked chip. Does that mean I can possibly change it?

Edit: Looking a little closer, I see that PL1 is not the only throttle reason. There is also the EDP Other on the very right side that often lights up when power throttling. I have no idea what that means and I assumed it didn’t matter, but since HW info confirms my change of power limits, it looks like dell has some junk that automatically kicks on the throttle.

 

All depends on how the Notebook manufacturer has put up firmware. If they haven't created good enough cooling, they will cripple it.

Intel have give the Notebook manufacturers some headroom how to put the power limits. Intel recommend PL1 100w and 125w for PL2. But it seems none of the OEM's will follow this due they save $$$ on designing too weak cooling. Or too thin models.

FYI. Dell increased the power limits of the processor in the XPS. As long as the temperature is not too high, the short-term consumption can be as high as 78W (for 28 seconds) and then 56W for sustained load of up to at least 2.5min depending on load and temp.

 

Correct. But...on my 9550 & 9560, boosting PL1 & PL2 limits/turbo-time via buggy Intel XTU did reduce throttling (probably because it bypassed some Dell throttling schemes). The difference was not earth-shattering but it eliminated all CPU throttling for my daily uses so was good for me. This only works because the Dell engineers did not bother to test their throttling schemes carefully.

Mind you, I also spent a lot of time undervolting, repasting, and tweaking to get these laptops to perform close to marketing specs lol.

Spot on!​

 

You can bypass some of the throttling schemes with a decent undervolt/colder hardware. And if the rest of the component on MB have better temp it will help as well.

 

I don’t get why dell won’t allow Intel XTU to override their trash thermal limits. If it wern’t for the slightly lackluster screen featured on the MSI GS65 it would be a no brainer for me.