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

Tony, I would choose to wait a month, maybe two. follow this thread to see how many people's machines are still dying (hopefully none) and see whether they will be getting a revised version from RMA.
That why I'm following this thread myself, I had the 51M in the basket already but thanks to this thread I choose to wait as I don't want a barebone Clevo.

 

Definitely a much safer route. Only seen about 2, maybe 3 defective 2070s, one of which was mine.

As for the 9700k/9900k for gaming, go with the 9700k. The difference in fps between the two is within a few fps.

The temps on the 9700k will be much lower as well.

 

9900k is not worth it for gaming as it adds heat, it may even be a detriment as you could possibly run the 9700k at a slightly higher clock than the 9900k due to less heat. But silicon lottery as well.

 

I feel good with my 9900k, 2000 points on cinebench r15 with stock clocks, -100mV, 80 ° C max with performance mode fans

 

I pulled 197w with afterburner, does that mean I have bios 200W?

 

i can pull 200+ watts with my 180 watt vbios on my gtx 1080. Compare your current vbios number to the 200watt vbios number. Only then you will find out 100% if you have the 200watt bios version.

 

My honest opinion: Take the 9900K and delid it. It will make it run much cooler and you will be happy for many years to come. 9700K doesn't have any hyper threading in it and the performance difference is going to be noticeable when you do a cpu intensive workload on it. And the 2080 is safe if you get the revision 2 card (but whether you'll get a true revision 2 card or not depends upon luck at this point. You could make sure with Dell that the GPU you're getting is a revision 2 card that has upgraded components).

I personally felt that the GT75 speakers definitely got a bit louder than the P870DM's and were definitely more crisp & clear. The low end was insane! P870DM's speakers are excellent too, no doubt about it.

 

https://www.tomshardware.com/news/asus-rog-mothership-zephyrus-price-specs-release-date,38312.html

This could be some serious competition for AW if this is released. Factory OC and LM applied is the way to go, and really caters gaming. Now looking forward to how the innards look, but the 51m still looks like the better choice.

 

If my A51-M dies I will get the money back and wait on this.

 

Agreed... They almost had something that caught my eye, until one sees it is a BGA tragedy - a 45W CPU that has slower base clocks than its predecessor. I think from a performance standpoint, the Area 51m is safe.

 

Reinstalling the 200W VBIOS would be fine.

 

Sometimes when i boot my laptop for the first time and i want to start a game my graphics card doesnt run at full speed, my fps would then be like
half from what i should get.

restarting the laptop would then fix it but anyone know what could cause this ?

 

Got a new Alienware Area-51M with i9 9900K & RTX 2080 today. Everything is working great, even with the latest versions of System BIOS & VBIOS installed. Something I can't figure out though is how do I switch to integrated graphics? I was hoping there would be a combination of keys to press on the keyboard as a shortcut, but I can't see any.

 

You can't. The screen is connected directly to the GPU.

 

Does this seem to happen if you boot the laptop before connecting a charger/s?

 

I think you only can use the thunderbolt connection to use intergraded grafix.

 

For reasons that I will not go into , I need to decide between the Alienware area 51 2070/9700 and the asus g703 with the 8950 and rtx2080. Could someone offer their opinion on the performance difference between these 2 rigs? I'm having trouble finding gaming benchmarks on the i7 2070 a51.

 

Well, the Alien will easily wreck the G703...no contest. You have 2 extra physical cores and much higher clock across all cores.

 

Sorry, I thought it was the 2080 A51M, my mistake!
So far I wouldn't buy the Alienware at the moment, even the 2070 version occasionaly dies.

 

We have like 3 years of warranty in europe so i'm not that afraid that it blows up. I have a backup pc for when it does. Been gaming hours a day on it and so far so good

 

Dell gives 3 year in the EU?

 

Pls What’s the vbios no for 200watss ?

 

It is installed on top of a Bios version 1.3.0 or higher. You are totally safe if you re-install it even if it was already installed. Download from here and install: https://downloads.dell.com/FOLDER05500081M/1/RTX2080_200W_VBIOS.exe

 

They give 1 year on site support but european law dictates that they have to give atleast 3. You won't get the next day help then but you'll have to send it in yourself

 

my vbios stock according to GPU-Z is: 90.04.3F.00.0E. every time i try to install vbios 200w it gives me this error: falcon reset control is disallowed! (PLM=0x000000EF)

 

Nice, I thought it was 2 years in total. But 3 is even better.

 

What's your bios version? It should be higher than 1.3, the latest is 1.3.2. Check that first before trying vbios upgrade.

 

my bios is 1.3.2

 

Question for ya fellas: With my Dell Rewards money I was going to buy a spare set of power bricks. There is no harm in buying two 330W adapters and using both of them together instead of the 180/330 combo, correct?

 

no problem according to frank azor, I will buy it too soon, the 180w heats up a little compared to the 330w

 

Shouldn't hurt anything. The laptop will only pull as much power as it is designed to take, you won't overload it. Plenty of people have bought larger adapters then what their laptops came with to get rid of battery drain or higher overclocking and no one has ever had a problem. I used my 330w adapter from my R4 on the M15 that came with a 180w brick and had no issues.

You will be fine and if you had any battery drain issues while under load that will probably get rid of them as well. The largest problem will be carrying around two 330w bricks haha.

 

Okay guys, help me out here.
Initially I tried downvolting the GPU as it was always around 81-82 degrees all the time when playing the heaviest games. Even on Overwatch with ultra setting on my new monitor which is 1440p 144hz (on the previous 4k 60hz it wouldn't reach (get close to) 100% GPU usage).
My mistake was, I looked up and found a guy who also downvolted his RTX 2080 on the area 51M, and did exactly what he did, getting virtually no improvement (later finding out that the I was most of the time below the frequencies I downvolted).

So I gave up on it. But hearing more about Area 51M's dying where almost always the cause was the RTX 2080. I got anxious and tried looking into downvolting again.
I properly monitored my GPU when playing games and checked what frequencies are used most often, as it is SO variable, due to thermals / power limits and the insane built in Boost clock frequencies.

So I ended up doing this. Seeing that the GPU clock frequency is 1545, I thought, Okay so during gaming, it will be at least this clock frequency. The boost frequency is 1725. But in reality. It boosts up to around 1950 (it is configured to go up to 1965, but it power throttles).
So when booting up a real game, and the temps kick in. The boost frequency is more between 1790 - 1880 Mhz.

But just to make sure, knowing that the default frequency is 1545, and also the frequency/voltage curve starts at 1320Mhz/700mv I tought heck, during non-gaming use, it never even reaches 1320Mhz, so I moved every single node on the frequency curve up with +75Mhz. AND to be stricktly below 1950Mhz. So I either increased with 75Mhz, or reached 1950Mhz.

On Overwatch, the temperatures Dropped about 8 degrees. which is amazing. Then I booted up Metro exodus, which only dropped like 3 degrees. And I also noticed Very minor visual glitches in metro exodus. So I wanted to lower everything to 65 instead of 75, but then I noticed, that GPU-z shows my default and Boost frequencies higher than the default ones as if I overclocked my GPU, Is that normal? Even with the slightest change on the Voltage/frequency curve, I notice on GPU-Z that my frequencies go up with like 1 Mhz. Although the frequency offset slider on MSI afterburner was still at 0.

Once again is this normal? Also apart from noticing the single minor glitch on Metro exodus, is there another/better game/benchmark I can use to test the stability of the downvolt?

Also seeing these results made me a little happy, I think if this properly succeeds, I will also downvolt the CPU after all, just for good measures. Less heat is better after all. And I don't see real harm in downvolting.

Edit: I figured it out something. When there enough load to make the GPU utilization reach 100%, The GPU goes to power throttling. Causing the max mv to drop from around 1030-1050 to around the 925-975 ish mv.

So due to power throttling, there essentially the boost frequency drops. Causing also the mv to drop.

Because there is a throttling that occurs, when I downvolt my GPU, it simply still reaches the same mv but this time around with higher boost frequencies. Because now due to the undervolt, there is more headroom before power throttling kicks in. So that headroom is used to simply reach higher boost frequencies. Essentially causing the voltages while gaming to be still around 925-975, only this time the clock frequencies are higher.

So basically have to figure out around what mv the power throttling occurs during max load. And I have to create a custom Voltage/Frequency curve that has 2 flat areas instead of the only one that is at the very end.

One for when a game is not utilizing the gpu at max load. Basically when the gpu is not throttling. Basically when a game reaches more than 144fps (which is the max of my monitor) (but since I have v-sync enabled, it doesn't calculate more than 144hz, essentially lowering gpu load)

And one form when a game is under power throttling. Basically when the GPU is at 100% utilization. Basically when I am running games that cannot reach 144fps.

 

Ok received 2 good news today :

- Rammstein released a new song (k you don't care ^^)
- i called Dell to know what's up about my repair in their workshop and the support told me they decided to proceed to a swap... maybe changing to warranty from HID U.S to me (E.U) helped...

My problems now :

- get my components that i left in the computer back to me (Dell told me this may require to send the unit back to my house then send it back with the Dell swap)
- two parts that HID changed when i ordered (memory and Hdd) are missing because HID didn't sent me the original parts in Dell's order. Support told me this may not be a huge problem but i need to get this clear with Dell support so i sent a mail...

More updates next week

 

Keep us updated. And you definitely deserve a new system! All is well it seems. Cheers

Sent from my SM-G965F using Tapatalk

 

I think we need input from people who actually have tried this with the Area-51m. I hear it can be very specific about what adapters you are using depending on your configurations. It would not surprise me if a 2080 equipped Area-51m refused to do correct power draw unless it has a 180+330.

At worst, could just buy the extra brick and see how it works. If it doesn't, just return it! Then you get an extra 90 days to use your Dell Rewards. I usually use my Dell Rewards for warranties though, since their site rarely has good deals.

 

Thanks man !

Hope that's the end now !

I'll be letting my swapped unit run for about a month, if it hold on then i'll repasting it with Gelid GC Extreme (changed my mind to repaste it with Kryonaut from what i read here and put some pads From Thermal Grizzly on the hot parts...

 

All those of you rocking sub 80c on the 2080.... is that stock or after undervolt/repaste?

Stock I'm hitting 84c pretty consistently, and pretty much immediately. But I haven't done ANYTHING to it. I'll tinker this weekend, I guess, but I'd like a heads up on if it's a dud or not... everyone else seems to be commenting about how high that is.

 

The GPU dropping clocks and voltage are similar to XPS on CPU+GPU usage.

 

@Papusan @Mr. Fox @iunlock

I wanna do yet another repaste, do you suggest I get the new Coollaboratory Liquid Extreme or Thermal Grizzly Conductonaut? Or perhaps the new Coollaboratory MetalPads? Do those metal pads just sit between the CPU and heatsink with no need for any additional thermal paste?

 

Coollaboratory Liquid Extreme or Thermal Grizzly Conductonaut should be equal. I wouldn't go for the metal pads. They don't work very well.

 

What about the new Coollaboratory Liquid Extreme

This sounds promising but I am not sure about the performance:

 

As I said. Skould be equal as Conductonaut. I have stopped many years ago with Coollaboratory Liquid Ultra.

And Silicon lottery havent switched back, they continue with Grizzly https://siliconlottery.com/collections/sl/products/delid

Remember Coollaboratory needed add in the Extreme variant because many people stopped up use their Liquid Ultra.

 

Okay this curve seems to work pretty nice.
Basically at 900mv when power throttling
And at 975mv when not power throttling.

However there is an issue (first of all you can see a small increase in the final node of the 1st flat spot, that Is automatically done by MSI afterburner when I hit apply).

The issue is that Even though the First flat spot start at 900mv, I still get values such as 906mv instead of exactly 900mv. While 906mv shouldn't be possible to reach with that curve. As it is the one next node after the first 900mv node, and both are at 1890Mhz.

And then even though 900mv and 906mv are both on 1890Mhz, MSI afterburner Shows that I am running at 1845Mhz at 906mv.
Like WTH? How does it get to that frequency at that mv value?

Am I missing something?

But apart from that, when power throttling the mv is now around the 900ish mv, and when not power throttling it's around 975mv.

Which in rough lines seems to work, but still not quite. So any help regarding that?

Because thing is. on the 1st flat, I did a downvolt of 81mv.
Despite that, when I'm running at 906mv at 1845Mhz, I essentially only have a downvolt of like 20-30mv in reality, because it doesn't exactly follow my curve.

And I didn't want to further increase the 81mv downvolt. So any help on how to understand these values properly?

Edit: Okay I found out that the voltages are too low for the frequencies I chose, which is why I see 1845 instead if 1890.

Okay then. I will have to do lots of testing to find the perfect curve for this power throttling machine. And I hopefully I will get there.

Also it seems like the curve changes on its own depending on the current temperature. Completely moving down the higher the temperature goes

 

You are thermal throttling before power throttle on RTX. The standard 180W vbios is more than enough to reach the 2100hz overclock cap. Of course, lower the voltage will impose a "power throttle" if too aggressive for your clocks (if not just crash)

MSI afterburner will never be exact due to the nature of bost clocks. It will only be "around" your curve, never exactly your curve.

 

I'm installing all the drivers kindly provided by Ultra Male right now.

By the way I can confirm that BIOS was 1.3.2 out of the box...

 

Long time lurker here. I just wanted to share some results of my Area 51m.

Intel Core i9-9900K
16GB DDR4 2400Mhz
RTX 2080
1080p 144Hz Gsync
1TB 970 Evo

First of all I am very pleased with this machine. I received it about a month ago and did not really have any issues so far.

The only problem I have is that my CPU has some coil whine, which is kind of annoying with the fans switched off. But while gaming it's ok.

I thought about contacting Dell about this issue. Maybe they will replace the CPU. But as there is silicon lottery I might be worse off. So I would like to know how well my current CPU performs.

The system is totally stock. I am on Bios 1.3.2 and 190W vbios. Ambient temperature is about 21 degree Celsius. Laptop is placed on a KLIM™ Wind Laptop Cooler.

I got a good result in CB15 with an undervolt of -130mv @ 5Ghz and priority set to high in the task manager. But sadly it's not stable at all. One day I did 4 runs of CB in a row and it did not crash. But most of the time I get BOSD. The XTU stress test (5min) is also impossible to survive. With an undervolt of -95mv I can complete the stress test but not without thermal throttling. Fan profile: Fullspeed. Seems my CPU needs about 1.29v to work @5Ghz.

CB15 -130mv:
https://www.imghub.net/u/Raul7/img/306641bb62

CB15 -95mv:
https://www.imghub.net/u/Raul7/img/6f80686020

For the GPU I used the OC Scanner in the MSI Afterburner Tool. I got an average OC of 135Mhz on the core and applied the suggested curve. Memory is set to +400Mhz.

I am gaming @4.5ghz with an undervolt of -125mv with good temps. HWinfo shows 1h 40min of Battlefield 1 on ultra settings. Fan profile: Balanced.

https://www.imghub.net/u/Raul7/img/58af9ab051

Are these acceptable values?

 

Thanks for the advice. I'm going to repaste it tomorrow using the X method just to be sure (for my own piece of mind) and then I will try the undervolt method. Seems a shame that we have to undervolt, etc, just to get acceptable temps from an expensive pre-built laptop, but I guess that is the point of an enthusiast machine. Oh well. At least the paste did seem to settle a bit into lower temps, but I still don't like hitting 87c cpu when playing something like total war. So, I suppose undervolting will be my next stop after repaste. Does 79c average (sometimes up to 84c) for the gpu sound good? Or does it need undervolting as well?

 

Sorry for asking but what fan profile are you on? On balanced in my 51m my RTX 2080 spikes to 80c after a few minutes and then my fans ramp up and it goes back down to 70c. At 100% load and with performance profile my 2080 never gets any hotter than 74c and is usually at 70c to 72c. All i did was repaste with kryonaut and to be honest the stock dell paste performed more or less the same.

I wonder if since the 9900k runs soo hot and the heatsink is a combined heatsink if the gpu is being roasted by your cpu. 84c is perfectly fine all things considered. It likely wouldn't impact longevity for your card. I have an 8700k and it is usually at 70c or lower in games at 5ghz.

 

Post your AB curve graph, please.
It will also help if you post the stock graph (or even the one with a manual +mv overclock offset without touching the graph) so we can see what changes you made.

Usually you "undervolt" (well, underclock) by finding your target voltage and frequency point on the graph that you want and then flattening everything to the right of it