*OFFICIAL* Alienware m15 Owner's Lounge

Is this better than the legendary Notepal U3?: https://www.amazon.com/IETS-Cooling-Flexible-Colorful-Adjustable/dp/B07WVK2ZZ9

@Papusan @Mr. Fox @iunlock

 

Nope. Wasted money. All notebook coolers sucks. A mod U3 is the only one worth to have.

This one is with a awful design. Plastic fantastic. You can't force air into the bottom grills with this.

 

I really cannot imagine that working well. It seems like a terrible design to me. I doubt that it has fans that are strong enough to move sufficient air to be more beneficial than simply elevating the back end of the chassis. The air outlets also seem to be too small and those thick plastic grilles would block a significant amount of what little air the fans are capable of moving.

 

With use of this sealed box, this will most likely hamper the notebook fans to breath. The two weak 5v fan won't cover up for this. The coolers design is awful and in function more like "as you yourself try to breathe through a thin straw". We all know how that goes Or more as you have put your notebook in a small drawer and hope it get enough air. E-waste!

 

Yeah, this is terrible.

1. This sealed design doesn't work - I actually tried this approach with my 5 fan cooler which is demonstrably very helpful unmodified, by sealing off the edges etc to force all air into the underside of the laptop. This significantly increased the temps. Not an aerodynamics expert so not clear why.

2. The fans look tiny
3. Can't see airflow being quoted
4. There is a filter which will further reduce airflow - a better approach would be to ensure there is no dust in the air using vacuum cleaners and external air purifiers

 

That's why i always build for the German AW-Forum decent ones. For the best possible cooling performance.

 

I've been using one of thesenfor the past 5 years:

https://www.amazon.com/Laptop-Cooling-EletecPro-12-17-3-Laptops/dp/B07GR5LMLM

Cheap, quiet, reliable, 75cfm, and the fans do bring the temps down by a further couple of C vs just raising the back.

 

Of course, a cooler is better than none, but the two fans below can't really cool the m15.

With raising the back you've better temps on CPU and GPU but worse temps on SSDs, PCH, RAM, Mobo, Keyboard etc. Tested several times on the R4 and R5.



Translation
Flach auf dem Tisch = flat on table
Mit Erhöhung 2,5cm = with raising back 2,5cm
Mit U3-Kühler = with U3-Cooler

 

Half of the fans just blow into the plastic bottom cover on most laptops.


Edit. The only ones who will benefit from fans at the bottom it is the Msi GT72/73/76 series.

 

Remember that the bottom part of the chassis also gets warm so part of the cooling effect is probably due to the cooling of the chassis. Also, it's not clear that the bottom fans don't contribute to the airflow inside the laptop at all. I *think* I did some benchmarking with the bottom fans off a while back and found higher temps.

 

It's not clear why raising the back would worsen the temps of components located in the lower parts of the laptop. Warm air should travel up and away from those components. It's a red herring anyway, since CPU temp is the only real issue in the m15.

 

BTW the bottom cover of the m15 is made of a metal alloy, not plastic. I know you hate AW, but be fair

 

I doubt it. I have 3x 3000 RPM High Pressure radiator fans on my cooler mods. They do almost nothing for small laptops with almost no grills at bottom. Metal or not doesn't matter.

 

I saw the results with my own eyes and posted them in this thread. It's true that the majority (about 66%) of temp reduction was due to the raising of the back. I would think though that, as is often the case, there are diminishing returns, and would bet a fair amount that none of the fancy custom coolers do significantly better than the £20 model I'm using. As with the custom PCH mod discussed in this thread earlier, crafting some custom coolers is an almost complete waste of time (in practical sense, it can still be a cool (!) hobby) as far as this laptop is concerned.

 

Some notebooks will benefits from PCH mods. Because some start throttle down speed or being unstable once PCH temp start crawling above 90 C. All depends on what throttle algorithms and temp sensors the manufacturers enjoy put into firmware and MB. It's an disgusting business. They will do almost everything to stop high RMA numbers/cut tech support costs.

 

For serious benchmarking: 3x Industrial EBM-Papst blower fan (from 6V up to 24V). This cools your notebook so damn good that it flies away.

 

Def. Some AW17s benefitred from PCH mods. No point if it runs at 50-60C though.

 

We' ve discussed this topic here a few times.

Dell knows the issue with the hot PCH (80°C+) in their thin notebooks. That's why they started to cool it out of box with the R2-lineup.

Heatsink m15 r2 with PCH cooling:

 

Yes., with the clear result being that PCH cooling is not required for the m15 R1 in normal conditions, in the sense that it has no impact on benchmark results.

 

Welp my 2 week old m15 r1 is running at 101°C+. Touching the power button makes me jump away with how hot it gets. Contacted support they ran some remote tests and they want to replace all the cooling. So now my brand new laptop is basically a refurb.
Since it’s only been 2 weeks just going to return it.

I have not touched/repasted the laptop. Everything is still stock

 

I don't remember the exact number, I am pretty sure it was something like 91-92C that my PCH would hit and then cause stuttering in games after 1-2hours or so. Now the PCH pretty much is consistent with CPU temp, low 80s after long sessions. It would only stutter a for a few seconds before it went away and came back in 30min or so. But it was def annoying in FPS games. Rendering and casual games nobody would likely care and its probably why Dell never cared. But its good that they revised their HSF even a little. Even if they are late af compared to others. I think everyone would have rather they fixed the alignment of the HSF in general, cuz that would help out the CPU, GPU and PCH at the same time.

I bought some copper RAM heatsinks before all the corona bs started and I really wanted to try them out to see if it had any positive effect on the ambient notebook temp with CPU/GPU maxes (I know RAM doesn't get crazy hot). But I went back to using the Dell 2666Mhz RAM since it seemed like mixing my 2400Mhz with a 2666Mhz was causing stuttering and lower fps than I liked since moving to a 240Hz. Anyone have any idea of when DDR5 might come to notebooks? I can't justify buying a 2nd 2400Mhz stick when I don't even use 16GB fully. I guess it can't be too long right, if notebooks are now being sold with Ryzen 4000 CPU, which afaik is basically Ryzen 3000 series.

Yeah I feel like you need one of those crazy modded coolers with multiple Noctua's to see a big difference vs raising the back. Sk0bold def showed good comparison's with his, but his is by no means cheap to make. I think there is a huge difference between a 60$ pre-made cooler and a 150$ modded laptop pad. If I didn't have a AGA which made the laptop pretty much unnoticeable when gaming on my TV, I'd probably make one. But then again, these laptops are kinda loud even at 60C

 

When I was repasting and tuning this laptop last summer I took some photos with a thermal camera of the bottom of the laptop without the cover. RAM, NVME and PCH were in that order the (by far) hottest things there. I added some thick pads to all of them so they touch the bottom cover. I don't know how much it improved the situation as I cannot take the photo of the inside with the cover on, but the cover is clearly hotter in the places where the pads touch it, so I suppose some thermal transferring is happening. All of this was done just because I like expending time modding my laptop, there was no impact at all in benchmarks or anything else.

About coolers, I tried custom cooling with a big 24V very high pressure 120cm, and also with some high pressure Noctua I had around, the difference compared to just rising the back was less than 5C under load. I think that with the fan grill the m15 r1 has, you cannot do much more than what the existing fans are doing. Of course if you go medieval with your laptop and "improve" the grill you could probably attach an air-conditioner to it, but we are talking about thin laptops

For desktop-like usage I also have a modded AGA (with a corsair PSU and a Noctua fan. That's by far the best (only?) advantage of having an Alienware. My 1080 Ti in it runs pretty cool whatever I do and when gaming the laptop stays under 75C.

 

That's very strange, perhaps I won the PCH silicon lottery, maybe it's the magic of my cooler, maybe you tested on Quiet fan setting and I tested on High Performance Could be a difference in ambient temps. I tested this in autumn, so I guess ambient was around 21C - could be somewhere in the original post. The stuttering could have been caused by something else too, for example the mixed RAM you mentioned (never a good idea).

I don't really have much to add on the m15 R1 PCH mod topic: 59C before, 48C after during benchmarking on the aforementioned cooler and in High Performance fan mode:

http://forum.notebookreview.com/threads/official-alienware-m15-owners-lounge.825553/page-336

 

I have a question for the group, how long does it take to get to the UEFI boot? Please include bios version. I have BIOS 2.3.0 and it takes 5s to get the ALIENWARE white logo and another 15 seconds before starting the UEFI loader (in my case CLOVER). I have a Thinkpad W520 ultra tuned with coreboot and a Ivy Bridge and it takes 0s to start GRUB. Whenever I use it I hate my Alienware.

The strangest thing is that if click F12 to select the boot device everything goes much faster, to the point that even with the time it takes to press the corresponding keys and select the right device I can boot in less than 10s.

I would like to compare with other BIOS versions, maybe that's why you cannot longer get 2.3.0 from Alienware support site :/

 

I am using 2.2.1 and i got the same time durations as you

 

That is a good point, that the AW case might be hindering it a lot. It's not like many of us would test it without the bottom case cuz then you may as well buy a desktop if you can't easily take it around. I think I've read that RAM might perform better when hot, like SSD storage? Idk for sure and that would be hard to test.

Well benchmarks are short. Cinebench only takes a couple minutes. Heaven takes maybe 10min. FF15 benchmarks take about 10min or less. But gaming for hours is what really raises the PCH temp for me. The heatsink is already glued so I can't make a comparison example of the different benchmarks VS gaming. But there is easily a 20C or more difference

 

We discussed the topic back then. The reason for this mod isn't for benchmarking.

Everyone should decide for themselves and I'm not forcing anyone. My PCH ran hot, so I gave some examples of how I solved that. Just wanted to help.

I've added some new informations about the m15 with the 8950HK and RTX 2080MQ. It's a bit different to the GTX device.

Repaste & Repad m15 (translated via google)

Modding & Testing m15 (translated via google)

I've two m15's. One with 8750H + GTX 1060 and the other with 8950HK and RTX 2080MQ. The PCH on the stronger one runs much hotter, even with a heatsink and cooler. Under perfect conditions = room ambient 20-22°C, LM, UV -165mV, cooler with noctua fans, all fans 100% speed, the PCH temps goes while gaming approximately 60°C. Flat on table without the cooler the PCH temps goes easily over 85°C.

 

To add more data to the PCH discussion, (or maybe noise), I've found that the temps depend a lot of what kind of I/O you have. For example, in my case, gaming with a Bluetooth xbox gamepad would increase the temp of the PCH by up to 15C compared to gaming with the same xbox gamepad connected over USB.

 

That is insane, I never would have thought of that. What else does the PCH get stressed from? Does the PCH have anything to do with NVME storage or storage in general? Or is it mainly for devices? I wonder if I can narrow it down. Because I have a 8-port Anker USB 3.0 HUB that I use daily, but sometimes I also forget to turn off the 2x 3TB HDD plugged into it

 

Latest LAN/WLAN/Bluetooth Drivers 13-Apr-20

Spoiler: LAN/WLAN/Bluetooth Drivers versions

Killer E2X Ethernet Driver v9.0.0.50

Killer E2K Ethernet Driver v10.038.1118.2019

Killer E3000 Ethernet Driver v10.039.212.2020

Killer E3100 Ethernet Driver v1.0.0.13

Killer 1550/1650/Intel WLAN Drivers v21.80.0.4

Killer 1202 WLAN Driver v10.0.0.355

Killer 1435-1535 WLAN Driver v12.0.0.938

Killer 1525 WLAN Driver v12.0.0.448

Intel Wireless Bluetooth Driver v21.80.0.3

Realtek LAN Driver v10.35.510.2019

Don't forget to unblock the ZIP file before unzipping it otherwise all the extracted files will have a BLOCKED status since this is a file you downloaded from the internet and Windows puts a BLOCKED status on any downloaded file which causes incomplete or erroneous installs. For how to do this, see:
"Are you sure you want to run this file?" [Yes or No]

 

I have an update with my problem recently i have typed. I was having problem with my laptop cause of faulty heatsink and faulty screen but technical service bring me only motherboard and screen. So screen was replaced without any problem but new motherboard was als faulty. So i asked them to replace my laptop. I don't know how they manage to bring me these faulty pieces.

 

same thing happened to me. I repasted and I must have fried the board and other components because the replacement motherboard they brought didn't fix the issue.

They were going to do something similar and bring another motherboard and screen but since there was a very long delay on the parts they escalated the situation and replaced the laptop. Hopefully they can do the same for you.

My replacement has been so much better. Was getting mid to high 90's in AIDA64 tests and gaming but undervolting did wonders without having to repaste yet. It may have reached high 80's once or twice but it def hovers at mid 70's to low 80's with the back propped up so I'm happy. I have the paste ready to go if it starts to get hotter.

 

I hope i can see those days as you right now my experience is horrible. Thank you for sharing your time to tell me about this.

 

BTW anyone seeing any undervolting issues with BIOS 2.4.1 or otherwise? There is some noise about this, but all is fine for me at least.

 

Are you using Throttlestop for your undervolt? I was wondering how throttlestop coexists the awcc. Awcc changes the pl1 and pl2 limits based on which thermal setting you have, but I thought throttlestop took control of those too, so what happens?

I've seen a few instances (on the interwebs) where pple had their computers royally fubar'ed after using throttlestop,,,, but I really want to try it as I usually have 15 minute averages of mid 80's to mid 90's across all cores with TONS of thermal throttling, but it really doesn't seem to affect game play, I only notice it when I look at the HWinfo logs so I feel like I should leave well enough alone

 

Every person who has this computer and lots of review shows that this computer runs hot. All we can do uv repasting and using extra cooler. Thanks for still using 14nm in 2020 Intel!

Cause of performance drop, i reverted to 2.2.1 and things are fine on this one.

 

Thinking of picking up a AW m17 r1

Have stuck with AW for a long time. Looking at other brands, but still wind up back to wanting an AW.

However the refresh seems to be plagued with overheating issues.

Would an undervolt/ repaste with thermal grizzly (now liquid metal) make a difference? Or is the fault due to faulty design?

The model I'm looking at is with the the i7 8750 & RTX 2070 MQ.

Is it worth a buy and a would tha AGA make a difference down the line? Or is the CPU not good enough for something like a 1080ti?

Thanks for your time & input.

 

They have improved the fans and heat sink every gen. So it has been already known that r1 heats a lot. And if you get 2070 mq it will be even difficult. My suggestion can be get 2060 model if you wanna go with r1 or r2. Or wait for r3 series to see how will they perform under newly designed cooling with new gen. hardwares.

 

Is the motherboard going to be inverted with the r3, and a soldered in ram?

If so it's a deal breaker for me.

Another option is the older Acer Triton 700 with a 1080 MQ.

Weird that the 1080MQ out performs or has similar performance to a 2070MQ

 

Yes, inverted mobo, soldered ram an wifi.

 

Yikes. No go for me.

Does repasting and undervolting make it bearable?

 

If you use the 2.2.1 BIOS and undervolt the CPU + GPU, its pretty reasonable. Considering its such a thin + powerful notebook, its effective. However, it could be way better if they designed the HSF better. depending on the task (YouTube + chrome typing this + steam, discord, b.net, few others) my laptop fans are inaudible. Idk if they aren't running or if they just can't be heard. However, its lowest speed can be heard, though its not annoying. And it can sound loud when doing something that runs the CPU + GPU both high, something like a emulator w/4K+ resolution. However normal gaming its pretty reasonable, it never hits 'max' fan speed and I never hear the fans with my DT990 headphones (they're kinda open back, kinda not, thats why I gave the name).

Basically, the r1 imo is a great notebook. Way more enjoyable than the 15R3 I had

 

Had a 15r3 go dead on me and am getting a M15r1 shipped out now - not a bad upgrade as I am coming from 7820hk and 1070 to 9750h and 2060 (they wouldn't budge) and from what I gather a much better chassis.

I'm gathering up materials for the required mods and am wondering if there is a post that compiles everything somewhere.

I think the recommended stuff is:
PCH heatsink (3mm ssd trimmed to fit)
LM repaste obviously
SSD heatsinks on SSD - I'll use pads and copper shims to make contact with case bottom.

I saw a 17R5 POST here showing a load of ssd heatsinks all over the heat pipes - is that something that works here?
https://www.amazon.com/icepc-DIY-He...ords=ssd+heatsink&qid=1590605119&sr=8-21&th=1

I think these 1.5mm guys would work fine for that if there is room. Are there any other recommended mods? On the older gen we could bend some of the mounting tabs for the heatsink to increase mounting force.

 

I would also add like, some cheap af thermal paste to make a barrier for the LM. Or if you have leftover paste from like a aftermarket HSF (like some cheapo cooler master tiny tube).
The heatsinks on SSD work really well for me, even when using the silicone rubebr bands, I just bought new RAM and decided to feel the silicone bands and they feel just as squishy as the day I applied them in like, December/November. Though I did have to take the warranty sticker off my Samsung NVME for it. And from what I gathered, the memory shouldn't be 'cooled' but the memory controller should be. Memory performs better when hotter and the memory controller performs better when cooler.
As for the heatsinks all over the heatsink, I am pretty sure that doesn't do much besides delay how long the laptop takes to heat up. Which in a way can be good, but if someone did some tests that showed some good benefits I'd love to see it to be proved wrong. I probably wouldn't do it but it would be cool to see comparisons.
If you want to waste money on heatsinks on heatsink, I'd recommend @Sk0b0ld 's custom Laptop cooler. Its expensive for a laptop cooler, but a piece of metal with noctua fans attached to it will last way longer than some tiny mosfet heatsinks that will need new thermal tape/glue to reuse them elsewhere if beneficial. And Sk0b0ld does very good comparisons with his temp tests proving its very useful

 

Maybe you mean my mod from back in May 2018 in the R4 and later also in the R5.

For this kind of mod you've no space in the m15 R1.

Maybe my links could help you. They're all from the German AW-forum, but with the Google translator it's understandable.

m15 R1 repaste (LM) and repad: -Link-

m15 R1 modding, benchmarks etc. -Link-

All information about the cooler you'll find here: -Link-

 

So, lately I've noticed that when running battlefront 2, the cpu throttles down to 33.75 watts, and the clocks are down to about 3 GHz, but the average max core temp is 96C!! this is after playing for maybe 20 minutes, but its at 96C after playing for only 3 minutes. (throttles like crazy monitoring with HWinfo)

After playing long enough to have drained the battery somewhat such that it starts to power limit throttle down to 25 watts, the temps run about 85 to 87 average, but it still seems like alot for only 25 watts.

I don't remember it being this hot last year, and was wondering if this is a sign of the stock paste being in need of replacement.

Any comments?

I ended up buying some thermal grizzly cryonaught, and since found that many seem to have it pump out or bake off within a month or two. Can't re-past right now anyway because there is no isopropyl alcohol anywhere to buy.

 

If your stock paste didn't start out with temps that bad, I consider that lucky. But yeah, its pretty normal for Dell to say 100C or less is 'normal' for stock paste. You could always turn off turbo boost temporarily until you can repaste, it will still perform amazing for games that are multiple cores/threads even at 2.6Ghz. I noticed CoD2019 performs amazing even with HyperThreading off. Didn't test warzone with HyperThreading off though. Some games however do need that single/dual-core power, but thats crappy-ported games like FF13's.

 

So you think I should try a repaste, is it worth it?
Or just go right to xtu for an undervolt?

 

I think you should do both if you have the knowledge to do so. They are both easy. But you mentioned not having any way of getting off the old paste, so I would just undervolt for now until you can get some Isopropyl alcohol

 

Thanks illuMinniti, I am actually more nervous of undervolting since I've read some reports of peoples systems undergoing permanent (or at least difficult to figure out / undo) changes involving speed step, and other things after using throttle stop and then deciding not to use it anymore... ie deleting the ini file won't completely undo whatever throttlestop did.

Still nobody has answered my question about how throttlestop handles the pl1 and 2 levels since the AWCC also changes those on the fly.

However, even playing BFV without turbo still gives good fps (70-100), and temps remain in the mid 70's so this is a good option until I find isopropyl.