*OFFICIAL* Alienware Area-51M R1 Owner's Lounge

My 9900K + RTX 2080 unit arrived on the 16th May, with v1.3.2 installed. I've currently got the v1.5.4 system bios, and latest v1.0.0.5 180 W VBIOS installed but I'm considering downgrading the system bios to 1.5.0 due to seeing the GPU clock decreasing every so often in order to keep the GPU temperature under 75 C, very odd.

Dell really do need to sort this issue out soon. It's annoying to say the least. Other than that, it's a nice machine. Got 8TB of SSD storage, and 64 GB of RAM on this baby now, it's monstrously powerful.

 

He realised the 51m would be the last cashgrab, after that everything would become soldered, tanking the AW brand image (like Dell did not tarnish it already).
he's probably jumping ship to go to an competitor, seems like he couldn't steer Dell the way HE wanted anymore (probably). Reading his last message it looks like he isn't done with "gaming". Looking forward to the next brand that Azor will be giving up slowly to corporate devils.

 

O

Odd. I haven't faced this yet, but I might have been lucky up till now.
My temps in Overwatch full HD on the Ultra preset remain pretty stable with the stock cooling.

Now before getting into the debate wether it is either possible or impossible, these are my preconditions how I run my Area 51M:
I do not play on an external screen, that is what the Threadripper in the home office is for ;-). This is primarily a livingroom laptop, in case of the lazies, and not wanting to go sit in the office, but do some couch-gaming. Yea, me and the wife do that especially in the weekend; me on Overwatch, and she Elite Dangerous. Yeah, I am lucky ;-) (No kids, AMEN!)

I limited the OC on this thing to 4,5GHZ; after reading the forums I decided to go with that, no need to go all in for now. No changes on the GPU, other then the fact that the rig has the 1.5.4. bios and the latest vbios installed.
Fans are on balanced all the time, spooling up when gaming. For additional cooling I have a coolermaster notepal that I have had for ages, and that is still running fine. Never ran a gaming laptop without.

See below my temps after an hour or so of gaming:


Given, I know I am not running it on its full potential, but will let it free float now to 5.0GHz tops.
But based on the above I am either extremely lucky, or? Now this is a basic scenario, and if you use it kind of the same way ther is no need for a repaste yet, unless you plan to OC the heck out of it of course ;-)

Just my 2 cents.

 

With the latest repaste + repad (using the replacement heat sink + Panasonic pads for 2mm and FujiPoly Extreme for 1/1.5mm), it seems that my OC has further improved at a much lower undervolt! Now my 2080 is running at 1950 MHz at only 918mV as compared to previous 937mV! In game temps have dropped to max 71 degrees now with the new profile!

 

The throttling isn’t a bug. It’s a ”feature” to keep the 51M cool and safe. For Dell, there’s nothing to sort. Use older BIOS and VBIOS if you want best performance.

 

I hadn't noticed this before... Dell lists their 8GB Ram as 2666Mhz while it rates it 16GB sticks at 2400Mhz. Can someone who's bought the 16GB kits from Dell directly confirm this? Is it a 2666Mhz sticks running at 2400 or is it a 2400Mhz max capacity sodimms?

I had the 2 × 8Gb option and while it ran at 2400Mhz, it had the 2666 option in case Dell ever released a BIOS that supported it (highly unlikely now since even Azor has left the sinking ship that is AW) Now, I don't see a single reason why they'll ever a BIOS with an XMP profile. It might come in the Area 51M R2 but looking at this after Azor's release, I think we might already be at the EOL for Area 51m... so much for "upgradeable laptop", eh?



Sent from a Galaxy S9+

 

Dell picking what they have in the shelves. It’s confirmed that they offer both 2400 and 2666MHz sticks. Both run 2400MHz.

Remember Dell offer other notebook models which have support for faster than 2400MHz speed.

 

Well, Dell is going to replace my 51m.

Inifinite loop lel

 

I can somehow understand tricks like this in slim pancake laptops. But come on - it's freaking DTR - advertised as fastest and THE BEST in world. Turns out dell failed hard with their DGFF and now to prevent it all - throttle machine to death. And it's not some regular throttle - like lowering clocks by 200 mhz to be just below 75C. It's some weird throttle all way down to 400 mhz - impossible to play anything with "feature" like this. To think that i wanted my friend to get it as his new machine. Seems like the "look" is all it offers.

About azor - my guess is that he just failed and caused company to loose thousands or even millions of dollars on replacements. Probably he wanted to do one good thing and pushed for best performance software to backup his claims. Somehow he couldn't foreseen that dellaware and his engineers failed hard and GPU is in fact ticking bomb. Too bad for everyone unfortunately.

 

I have rebutted this type of messages so many times in this thread that I'm not even going to try again. I enjoy beating the hell out of my machine and it never disappoints me. If one installs 1.5.4 bios which is *OPTIONAL* by the way, and then complains instead of downgrading to older version of bios, it's his own fault. It's easy to talk a big talk but that seems to be the norm around here. Just big talk. People benching factory laptops and then complaining. If you want to bench, you make your machine ready for it first. If you bench without doing that, don't blame the manufacturer. As much as I disagree with a lot of Dell policies, I find it utterly ridiculous to see lazy bums whining because their laptop gave out while they were trying to bench it with factory cooling solution (Now don't give me the same "oh it died while playing a video" nonsense either because if somebody benched the hell out of the machine earlier without proper thermal upgrades, it WILL die ANYTIME in the future while performing petty tasks). Factory cooling solution is meant for REAL LIFE SCENARIOS ONLY! Benching is NOT a real life scenario! Full stop. Period. And a lot of "resellers" did a pretty "spectacular" job with thermal "upgrades" and they usually run a whole bunch of benchmarks before sending the laptop over to the buyer so if you want to blame somebody, blame the reseller, yourself (for not checking things before believing everything that the reseller told you) and THEN the manufacturer.

 

Btw my 51m is running the 1.5.4 and it does not suffer off brutal throttle.

 

Ooook !

Applyed some undervolt, 1960 Mhz full load and 0.900mv :

stabilized at 74/75°c @ full speed fans

My swaped fresh unit was equipped with 2x8 GB of DDR4 2666 indeed.

This is my setup right now :

Repad & Repaste Kryonaut on CPU and Gelid GC Extreme on GPU.

Undervolt @ -110mv on CPU (i9 9900k default multipliers) with Throttlestop
Undervolt @ 0.924mv on RTX2080 with MSI AfterBurner.

75°c at max on GPU on overwatch.
70ish at max on CPU on overwatch.

Bios 1.5.0. Capsule Update Bios Disabled.

And that's it...

A lot of work but nice results...

Will see in the future if Dell release a new bios... and if my laptop is still alive..

 

Why different pastes on both CPU and GPU?

Have you tried letting it go beyond 75 degrees celcius?

 

Yep, it cant go beyond http://forum.notebookreview.com/thr...-owners-lounge.826831/page-1312#post-10925412

 

Overwatch doesn't really stress the GPU enough, you should try playing Division 2 and Metro Exodus and I can almost guarantee you that your game will crash if you run it at 1960Mhz at .9mv (from my personal experience of course and your mileage may vary) If you do choose to try it, please post your results here.

Anyone knows what Eyesafe display technology is? My replacement laptop claims Eyesafe display while earlier one just had Tobii eyetracking and gsync. Dell's website says it will start shipping Summer 2019.



Sent from a Galaxy S9+

 

Just marketing mumbo-jumbo. It's the same.

 

It wasn't safe on 0.9 switch back to [email protected]... stress tested with 3dmark.

Is 3dmark enough ? Caus' I don't have division 2.

 

So I know when this thread was new some of us were talking about swapping out the base cpu for the 8700K/8086K. Has anyone tried that? I don't see there being a problem with that, but I could see Dell doing something odd like restricting it in software not letting you overclock or something like that.

 

I think the fact that it does not want to cross 75C proves that it throttles. Else you could've pump power through your chip letting it reach 100C easily (create a fan profile with the gpu fan off).

 

Spent the morning checking GPU temps. Bios 1.5.0 playing Doom 2016, I got the 2080 RTX to work at 76% and get up to 80 degrees on stock voltage and clock settings no throttling or issues with clock.

I have Panasonic thermal pads (1, 1.5, 3mm) in route and I have ordered Grizzly Kryonaut, and Conductonaut haven't decided which to use on CPU and Phobya nanogrease extreme for GPU.

I might have ordered too many thermal pads and I'm not sure yet as they have not yet arrived. I'll post and let you guys know if I ordered too many. I'm on the fence about wich themal paste to use on the CPU so your help on that is greatly appreciated! Thank you guys.

 

Also I keep seeing people claim that bios 1.5.4 forces the GPU to throttle down to unplayable levels (like the 400mhz he claimed). I am still on 1.5.4 with my 9700k/2080 and the worst I have seen so far is roughly a 150mhz downclock with a still very comfortable frame rate. I do plan on trying to revert, just too lazy as it has been working great for me so far.

 

You'll need a whole lot of 2mm as well. http://forum.notebookreview.com/thr...-owners-lounge.826831/page-1111#post-10911767

 

Looks like azor is AMD new chief gaming officer

 

yes it throttle after 75c, but smoothly, it does not cut performance brutaly. I just lost 7/8% of gpu performances, on a hot summer day.
In game its not noticeable.

 

100C? LMAO! 87 degrees is the throttle point SET by nVidia. If this guy wants to reach 100C then he might as well write his own drivers and perhaps make his own version of 2080 as well.

 

Tried some 3d mark Timespy extreme.

931mv seems to be stable.

80°C at max, with repaste and repad.

Still can see PWR limit while the test is running but it seems I can't avoid this even after proper and meticulous repad on the GC..

I'll let the test complete some other passes to be sure

 

making a GPU reach is quite easily even after it crosses the throttle point. You only need to edit some HEX values in the vbios rom file, and you're good to go not to mention that drivers are an interface for software developers to let their software work with the hardware else everything had to be coded in native c++ or any other assembly language for some simple programs. making hardware reach a certain point has to do with firmware, not drivers.

87C is where it get's throttled. OEMS have a different definition of throttling, and how hard something should be throttled. By unlocking the powerlimits (which cannot be done on the laptops cause Dell locked the vbios) you can still let the card reach it's 100C to perform an thermal shutdown. After that you can see (and analyze) how much Mhz you're going down for every Mhz, how much voltage get's throttled etc etc reverse-engineering the algorithm that trottles the card

 

Easier said than done. And, BOTH firmware and drivers are involved. Things are not as generic and "Black and white" as you make it sound like. It's not 1998 anymore.

 

20 minutes of 3Dmark TimeSpy Extreme Stress test : 84°c (never reached higher temps) no framerate drop no frequency drop.

Good boï ^^

 

@S.K

Hi, your posts are very informative. Thanx for so much feedback, even though I do not own A51M, I can apply some of this down to earth knowledge to my own laptop.

X
X
X

I am very happy user of Gelid pads, but would like, please, some feedback on the Panasonic 13 Wm/k pads if possible. (before I possibly order)

My main concern is how well they compress and if they can be torn apart (say LOL, with hands).
For comparance, I have owned Dell stock pads, Alphacool 17 Wm/k and Gelids only.
With Gelids, when I am not careful, they tear apart very easy on heat-sink removal. My 'imaginary picture' about the properties of Panasonic pads is that they will never really break, nor can tear apart.
What is your opinion/experience so far with them?

Thanx for any info you can provide, very appreciated.

 

Hey bro, I will take Panasonic pads over Gelid ones every time because of the following reasons:
1: Higher thermal conductivity
2: Extremely high compressibility
3: No tearing as the pads remain REALLY soft even after multiple weeks of hardcore usage of the machine, sometimes without an aircon running in the room!
4: As soon as the pressure is removed from the pads, even if they are compressed, they gradually return to their original form which I REALLY like! This is something that completely sold me on these, and I prefer using these over FujiPoly even though FujiPoly Extreme has higher thermal conductivity but those get dry very fast and are not re-usable (in most cases). However, where I want a set and forget setup and already know which size fits where, I'd go with the FujiPoly Extreme for the final build, but unfortunately they don't offer pads thicker than 1.5mm.

 

Hey @S.K ,

Can I have your advice on this :

Is 84°c on GPU after 20mn 3Dmark TimeSpy Extreme stress test something relatively reliable ?

Parameters : Freq: 1950MHZ@930mv set on After Burner.

Many thanks

 

Hey brother, sounds like you finally got a reliable machine! This sounds plenty good to me! However, I'd use the computer for a few weeks as it is right now and then re-visit the profile to make some changes. After "settling down", sometimes you can get a higher overclock at a lower undervolt as compared to when you started for the first time. So performance is only going to get better from here onwards. My GPU used to do 1950 MHz sustained at 937mV but after my recent re-paste and repad with a new heat sink, I'm able to achieve the same frequency at 918mV haha! The only thing that I changed is the thermal pads on those 2 pesky little mosfets that are hidden underneath the beauty cover. Previously, I didn't have a 3mm lying around so I sandwiched 2mm and 1mm on top of each other to "get going" but this time, I had a Panasonic 3mm lying around which I used directly. What thickness pads have you used in your repad project? Did you follow this guide in my signature below?

One more thing: make sure you didn't get robbed when buying kryonaut! Too many fakes being sold around these days. I had a fake kryonaut batch that gave me similar max temps which I had to return to Amazon. I personally am thankful to @Mr. Fox for mentioning Phobya Nanogrease Extreme as that stuff just blew me away! I am not going to use anything other than that from now on (apart from Gelid GC Extreme as that's a consistent mid tear performer).

 

Well which System BIOS, and VBIOS do you personally recommend for the best performance? I’ve read that the System BIOS v1.5.0 is the best to use, but clueless as to what version of VBIOS to use.

 

1.3.2 gives the absolute best performance along with the 200W Vbios. Dell has removed it from their website but I have uploaded them on my google drive. You can download if you want. In order to install that bios, follow the guide mentioned in my signature below.

P.S., After installing it, make sure that you disable the capsule updates feature from within the Bios in order to stop Windows from installing new versions via Windows update.

 

Yes they do, but not for throttling. Drivers are involved in a such way that they can for example let the hardware calculate something until the hardware crashes. But just like you said, they cannot anymore cause it's 2019. Nowadays you can only let the software crash, not the hardware. An hardware crash usually results in a broken part, where as an software crash is most of the times, an bit that got flipped (due to cosmic rays) or an catch and release that failed to catch an exception (resulting in an software crash). both are hardware crashes, where software is the fix. In the event a GPU drivers tries to do something it should not, it simply crashes. When was the last time that you read that someone's GPU was stuck @100% and burned someone's house down because it kept going 100% without stopping ? never cause the firmware is the brain of the GPU just like the BIOS/UEFI is the brain of the CPU.
without both the GPU would not be able to interface with windows, and without drivers programs that run in windows would not be able to calculate/interface with an GPU.

That's why a CPU doesn't need drivers, it's interface are it's instruction set it supports, where as a videocard can only make a certain type of calculations that have to be programmed specifically for the GPU.

Nvidia tried to get it in datacenters and such, but a component that needs drivers, will not work better than something that's embedded within the hardware already.

That's why you need to keep your drivers of your gpu up to date, it's contains all the latest and greatest in terms of HOW the gpu interfaces with software. Game developers do not program GPU, they write software that the GPU can interpret due to Nvidia doing the driver side fixes. No (game)developer will now how an GPU works because the drivers blobs/binaries are supplied by Nvidia and just like we all know they like to keep tight lipped about their precious IP. but i do get your point.

 

So the re-useability is then excellent !!! COOL!!

I ask, because with Gelids, as soon as I remove heat-sink I always change for new pads (about once in 6 months), as they never come back to shape, and I do not want to risk 'very weak or no contact' on black Mosfets especially after my 1060 GPU is now flashed to max. 130W (from max. 78W stock).

On every re-pad, with Gelids, I have to custom fit/measure the gaps because I must re-balance the fragile heat-sink from scratch for good die fit first, then I add pads while keeping dies in contact with heat-sink.

Can a 1.5 mm Panasonic pad compress to 1.1 mm easy?

I ask all this !!! LOL I want to try a new product as an alternative to, so far best low-pressure-heatsink performing Gelids on my 13 r3, seems like with Panasonics I can have more tolerance freedom and re-usability and better thermal performance AND pad longetivity.


Thanx again !!!

 

I have seen 2mm Panasonic pad get compressed to lower than 1mm with no issues. Those Japanese folks really know how to make quality stuff!

 

Thanks !

Yes finally, after all those up & down in my personal story about this laptop...

Yes I did use your tutorial about repad (I even created a backup.... just in case..) : here's the first application : http://forum.notebookreview.com/goto/post?id=10924273#post-10924273 (before I had to do it again because of the failed MB/CG cable).

I'm using a syringue of Kryonaut i've been using with another computer that is still runing fine.

If I got any problem I'll be switching to the Gelid I got or the paste you recommended.

 

CPU doesn't need drivers? What do you say about intelppm.sys? As far as burned GPU example goes, if a driver wants to override firmware defined variables within the on-chip memory where firmware executes itself, it CAN. There are proprietary switches and instructions that are reserved for the manufacturer provided software only that CAN override the basic firmware on any device. How else do you think overclocking works? As far as the 2080 goes, the hard set thermal limit is 87 degrees. To heat it up beyond that and still making it run is simply out of question (as you probably mistakenly quoted in your previous message). And by the way, I've developed a lot of drivers back in the good old Windows 2000 and Windows XP days so I know a thing or 2 about how things run.

 

I ordered QTY 4 EYG-T7070A20A

 

I noticed that Dell factory heat sink comes with 1mm thermal pads on the VRMs on DGFF (that long strip on the right side of those 2 memory chips towards the GPU fan that covers those 6 tiny bricks). Once I tried to use a hardened up 1.5mm on there and it probably created z-height issues as the heat sink would not sit flat any longer (guessed from increased GPU temps) because there was no pliability in the thermal pads. In my most recent thermal pad job, I have put 1mm there instead of 1.5 and feel that I get a tighter connection to the dye this way. I'd push down hard on the heat sink after installation just to make sure that all the pads are "seated" properly because sometimes those tiny pressure springs can't press the heat sink hard enough for the pads to compress adequately.

As I previously mentioned that living at just 2 minutes walk from the sea can really destroy your computer in the long run as contacts and pins get corroded. So, I did some quick upgrades to my motherboard to prevent it from corrosive damage (after saying good bye to 3 of my machines within 2 months because of corrosive damage ). I used Starbrite Liquid Electrical Tape for covering the sensitive points where usually corrosion takes place, including RAM slot pins contact points on the motherboard. Also, installed 4 heat sinks on the primary and secondary inlet mosfet pairs.











This will improve thermal stability as well as longevity of the computer to some extent.

P.S., If you are planning to do such a mod on your machine, make sure that you DO NOT cover any surface mount resistors or capacitors with the tape because that will prevent them from radiating heat into the air which could lead to overheating of individual components as well as reduced stability of the system. And most importantly, make sure that no liquid tape goes inside the RAM slots or it'll cover the pins, making your RAM no longer work. And while you might try and clean that off, you could end up bending/damaging some of the pins inside which is no good.

 

I think you're confusing Kernel drivers with 3rd party drivers. A kernel driver could only be provided by Intel whereas an Nvidia driver could be re-versed engineered to create one ones "driver". Alot of forums where people share custom drivers.

Overclocking has nothing to do with an driver overrideding switches or anything. You're simply sending an instruction that can be executed by the hardware (undervolting, or overvolting locking the frequency etc) those are parameters that are open to tinker with. For example my 17 R5 has a 110w hardlimit set in the bios that no software can override, only a new firmware. There is no software that can trick the 110w limit.

The temperature of 87C is the Throttle temperature not the "TJunction" of the chip. The difference between temperature and for example my 110w limit is that mine is a simple bottleneck imposed by AW where as the 87C temperature question is about a throttlepoint, not a hard limit. Reaching 100C will shut down the chip itself where as when it hits it's throttle point it just removes performance in small steps until its under 87C. But sometimes conditions in a case does not alow for that to happen letting it float at a high temperature or agains the tt boundary. At forums there are people that have crossed the 87C "Limit" and getting the feedback that its "normal". So yeah..


I get where you're coming from, but things have changed in the years. Not trying to downplay your experience, but hardware and software has come an long way as you know.

 

I think you didn't read my message closely enough. Plus, this whole rant is rather useless because form factor and laws of Physics are a real thing that cannot be defied by comparing what AW did in this chassis compared to liquid nitrogen cooled desktop hardware. So, let's save it.

 

I thought so... try 1950Mhz around .935mv and run some 3dmark or Heaven benchmarks, I mostly use heaven as I find it more like a normal game stress.

You can get Division 2 for Free with your Tobii eyetracking software. Just check there, I'm sure the offer is still valid.

From personal experience, although while stress testing your 1950Mhz @ .935mV might pass flawlessly, gaming for 2 hours plus is a different scenario all together. It's in these scenario you can truly decide whether your undervolt is stable or not. For me, the best TESTS for my daily driver settings was playing Division 2, Metro Exodus and Far Cry 5.

For most people who say your "5Ghz on all core overclock" is running under 90C while benching, play Far Cry 5 for one hour and post your results here. Otherwise, I will not believe it to be true. This game pushes the CPU beautifully and I love it.

Only judge your most stable settings only after 2 hours of these aforementioned games, they're the real world tests.

Sent from a Galaxy S9+

 

Any updates on 2666 ram support?

Anyone annoyed that small movements on the trackpad cause snapback of the mouse cursor and are sometimes interpreted as false clicks? Any solution to this?

 

No Azor lied to us, there will be no support for 2666 MHz RAM despite the smaller brother m15 having it.

No, because:


PS: there is an Edit button you know?

 

No matter what I set I still get the snapback

Any specified reason as to why 2666 won't be supported?
My Hackintosh seems to be running at 2666 (as far as I can tell). It doesn't work in windows though

 

no specific reason other than the fact that Dell doesn't care. They only focus on big problems. Frank Azor promised me 2 months back that there will be a BIOS update to support 2666 MHz. RAM and they released 4 difference BIOS upgrades since then but nothing. Now that he's leaving Dell, I wouldn't even think about it anymore.

 

Whose replacing him? Or is his position being eliminated?