*OFFICIAL* Alienware Area-51M R2 Owner's Lounge

9th

 

I wouldn't get too concerned, it is unfortunately dependant on so many factors, availability of parts, your location, operational tempo etc. I ordered mine on the 4th and it shipped 14 business days later, now en route from China to the UK.

What I do suggest however is raise the issue if its longer than what was initially advertised to ship (UK site has had 9-11 buisness days for last few weeks). Also check the estimated shipping date via the orders page and shipping details. I highlighted it was longer than advertised on both counts by phone, stated I was thinking of cancelling and they actually apologised, offered to attempt to expedite and gave me a £100 refund (to follow after reciept of the system).

Update: Just checked my tracking info and its changed ETA from the 5 Aug to 30 Jul!!! Woo Hoo!

 

What a joke I'm on holidays till Monday :'(((((

Was due the 5th of August

 

Sorry to see that, I hope you can arrange someone to take delivery?

 

My neighbor is my saviour lolll
So far they've been amazing with producing the units so fast

 

If I'm having fit/pairing issues, I default to Nanogrease.

If I have a decent fit, I use Gelid.

Like you, I've found it to be overall better assuming a good pair.

 

Unless "moar coarz!" aren't needed depending on work flow.

If you're a gamer, there's no way I'm going with an AMD laptop.

Unless you're doing something very multi-threaded dependent with no cap, there's no way I'm going with an AMD laptop.

I prefer to see testing single through max cores so you can see the transition from Intel leading to AMD eventually pulling ahead based on sheer cores.

 

Sure, how many cores do games really need? Not many. While we are at it, are there any games that will materially benefit from 10900 vs 9900? Quite unlikely. And since you asked, as an occasional gamer I'm very happy with a 8750h 2070 Max-Q laptop running my favourite esports title at 300 fps+ lol

On the other hand, multi core applications (content creation, engineering, science, etc.) will tend to use all the cores available and, CB20 suggests will run about 20% faster on a 2000 euro 15" 3950X laptop than on the fastest, much more expensive, 17 inch Intel laptop... Good. More choice and thus power to the consumers.

 

I just got to know what RAM is on my laptop. It's the Kingston XCCT36-MIE / XMP-3200

 

Just as I thought - basically Kingston Value RAM CL22, according to the first hit on Google.

 

While the information may be somewhat useful in identifying the worst thermal pastes that should be avoided, thermal paste comparisons like that are often not very accurate in terms of predicting what someone can expect with laptops. Placing too much stock in reference to them can be very misleading. Comparisons like this are generally compiled using a desktop with a much better thermal solution and rarely ever any issues with fit or contact pressure.

There is tremendous variance in laptops in terms of fit and contact pressure. In fact, there can be significant variance in results even with one laptop. Every time you mount and remount the thermal module on a laptop there is at least a minor difference from one time to the next due to poor fit and sloppy tolerances that are to some extent necessary because they are not manufactured to aerospace-grade tolerances. I can't tell you how many times simply removing the heat sink and putting it back on again with no change in the thermal compound has produced a significant difference in temperatures on laptops.

 

I can pick many games that will bring your modest laptop to its knees.....easily at >stock 1080 and I do mean with great ease.

Your argument of 9900 vs 10900 supports my argument of core requirements. It extends beyond games.

As for the second part, that’s why I said:

“Unless you're doing something very multi-threaded dependent with no cap, there's no way I'm going with an AMD laptop.”

Everything you said fits into what I said. If I need a butt load of cores that cross that performance line of intel’s 8-10 core solution, AMD is the way to go.

I definitely agree that competition is good, choice is good and more power to consumers is good.

 

@etern4l @Spartan@HIDevolution @Papusan @Mr. Fox Can any of you please explain to me the issue here with RAM CAS latency and data rate? As I understand it the "absolute latency" calculated would mean the HyperX 2666Mhz CL15 would be superior to all its variants at 2400 (CL14), 2933 (CL17), 3200 (CL20) etc.

Is this actually the case? Seems a bit disappointing that the 2933 or 3200 XMP options with the 51M R2 are inferior to the 2666 HyperX I had in my 17 R5.....

Also why would anyone purchase RAM (or rather HyperX sell it) of a faster frequency if the overall speed was inferior?

My knowledge is amateur at best so any sage advice or thoughts would be very welcome.

Update: Having done a bit more research I guess the key here is generally speed trumps latency, so get the fastest RAM with the lowest latency essentially?

Tempting to buy some 3200 HyperX and hope it works. Would also be intetesting if it didnt and defaulted as designed to 2933 (CL 17) what the performance difference would be vs the Dell XMP 3200 @ CL21-22?

 

It really depends on what you are doing. Sometimes clock speed is more important than latency. But, 3200 CL22 is really terrible though. I am running my turdbook at 3000 CL16 and my desktop at 4000 CL16.

What's not clear to anyone at this point (and may never be) is whether the people at Alienware don't have a clue what they are doing (very possible) or there is something about the Area 51M R1 and R2 that is not engineered properly to handle it. If the BIOS doesn't allow access to memory timings and memory voltage, then nobody will ever have an answer to that question. All anyone will be able to do is be ticked off about it and take wild guesses what might be the deal. At the very least, such sloppy loose timings at such a low memory clock suggest that Dell is cutting corners with low-quality components just so they can say the product supports 3200 RAM, which should come as no surprise here.

Besides that, do we even know that it will run the 3200 RAM at that speed, or maybe they're just selling it with 3200 RAM installed because they know people are going to buy it? Didn't they sell the R1 with RAM rated at a clock speed faster than what they would allow it to run, or am I remembering that wrong?

 

To its knees? I doubt it, but well, do share at least one then lol About a year ago pretty much anything I tried ran fine in at 4K at max settings, not to mention 1080p. Although, even then I had to dial the settings down a bit @4K in one case. That's hardly taking the laptop to its knees. If I look at Fire Strike results people get with the latest Area 51M I see, wait for it, 30% higher scores. That's right 23000 vs 17500. That's from a laptop with a GPU that runs at twice the power, and has twice the CPU multicore processing capacity. Just lol, talk about diminishing returns.

You didn't quite get my 9900 vs 10900 argument, because it does not support your claims - all we say and agree to is that both of these CPUs are an overkill for gaming. If 10900, and in fact 9900K, are not fully utilised, then perhaps single CPU tests will be more indicative of CPU utility for gaming. Let's consider 3950x then. According to notebookcheck, it scores about the same in single CPU tests as 10900, which might be better representative of gaming performance.

Thanks for sharing you would always pick Intel for gaming, however, other people might be more rational and we need that in the world today..

For me a 3950X laptop is out for two reasons: no battery life, and no TB3, but I wouldn't be worried about gaming performance.

 

Actually, the relationship is fairly simple: the number of operations RAM can perform per second equals the clock speed / CL. You can compute this ratio for any particular RAM module to reasonably predict its performance. Dell consistently supplies RAM which HyperX outperforms by about 20% in terms of both the number of ops/second, and sure enough, memory benchmark results.

In terms of what that means for real life performance, I would think gamers are less affected by slower RAM. My laptop is batting in TimeSpy at about 15% over stock numbers posted in reviews, in part due to undervolting though.

The point though is you should be able to get just 8GB from Dell, get HyperX online, and still save money, so it's kind of irrational to order more than a minimum amount of this slow memory from Dell.

 

Why is it that the original Area 51m seems to be outperforming the R2 particularly in Firestrike? I thought the 2080 Super was quite a reasonable upgrade over vanilla.

 

Where did you see that?

 

Just looking through the top scores on the 3DMark web site. For Firestrike, the 51m R1 has a 28526 (which is admittedly a bit of an outlier), 26814, 25663, 25270, etc. The highest score for an R2 is 23668 at the moment.

 

My FireStrike Score on my Area-51m R1:

No Overclock:



Overclocked:

Fire Strike (Driver 442.23-BIOS 1.3.2) OC 20 Core-150 Mem



Fire Strike for the Area-51m R2 posted by @ratchetnclank



That IS an improvement. Remember, the RTX 2080 SUPER is only slightly better than than the RTX 2080 so you won't see any mindblowing improvements. As for the 28,000 score that you saw, I have no idea how legit those scores are

 

Many is from eGPU/Sli. This is real. And I can’t see much of an improvement with Super.

 

This could explain why:

"
Power Requirements

NVIDIA estimates that the RTX 2080 Super should consume a maximum of 250 W, which is 35 W higher than the company rates its GeForce GTX 2080."

If you cap the power at 180-200W then the whole upgrade is pretty much pointless.

 

I am looking at getting the A51m R2 as my first Alienware device. I'm glad to have looked up these forms to see that it is worth self upgrading the RAM for cheaper and better quality, so thanks to everyone here posting about their systems. I'm planning to do the same with the NVME drives and swap them for Firecuda sticks. Planning 2x2TB Fircuda + 1x4TB Sabrent, 64GB HyperX, 2080S with the 10900k on 4K (I currently use my machine more for Data Science work and 4K video editing than FPS gaming(not to say I won't occasionally)).

Does anyone in the UK know of the status of Dark Side of The Moon coloring? I've been holding off for it since launch and it just seems like it isn't coming. Took me 20 minutes on Dell chat for a dell sales agent to say "At this time we do not know e Lunar Light varient avaILABLE" ... which wasn't really helpful considering Lunar Light is the only option for the UK atm.

@ Spartan@HIDevolution I'm thinking of doing my first DIY de-lidding, so be great to hear whether it was worth it when you get yours. I am curious as to why HIDevolution offers liquid metal between the die and IHS, & the GPU and Heatsink, but not between the CPU IHS and Heatsink? Are both heat sync contacts not copper? (Sorry will be my first time looking at doing liquid metal). Was looking at using Conformal Coating to protect any nearby contacts.

 

A shame really as they allow the CPU to pull up to that but i guess they are afraid of burning cards again?

 

It's weird, particularly given that the GPU is usually much easier to cool.

 

Dark side of the moon is selectable only on 1080p in the UK as you said but for some reason the US can have it on 4K too. I don't see the reasoning for this.

 

I guess I'm just surprised the gap isn't bigger. My m17 R3, with an i9-10980HK that can barely manage 3800 in Cinebench R20, is pulling 22,435 in Fire Strike despite an abysmal physics score. I was thinking of returning it and picking up an Area 51m R2 instead but I'm wondering whether the gap for workloads other than pure heavy multithreaded CPU is all that great after all.

 

A gaming laptop is not only about getting high scores in synthetic benchmarks. It's about having a laptop that does what you want. My current laptop is an Area-51m R1 with a 9700K/RTX 2080. sure I don't get great Cinebench results lik the i9-9900K (I only get 3557 in CINEBENCH R20) but it can run any game or task I throw at it with breeze. Yeah so what if I have to wait 5 or 10 seconds more while rendering a video? It does everything I want just perfectly so be happy with your R3 if it's doing what you want and don't burn too much energy on chasing benchmarks.

Just my 2 cents. If you're wondering then why am I upgrading to the R2 then? simply because I need more storage and want a better screen so I thought I'd upgrade to get the screen I want and more storage/faster RAM then sell off the R1

 

Wondering if it is in theory possible to do a shunt mod on this mobile RTX 2080 Super. On my 2080 R1 I did not find any shunt resistors. See, I´m really trying hard to blow up my 51M

 

bsch3r =

 

Firestrike doesn't weight the cpu/physics very highly at all. It's all about the GPU performance.

Timespy seems to take it into account more.

 

They are there, and are 5mOhm as usual, just flip the card.

 

One or two of those are probably from me (if I remembered to upload them) using a 2080ti, Titan V or Titan RTX in the AGA. IIRC, the EVGA 2080ti XC Ultra scored about 28k and the Titan RTX was just over 30k. 3DMark doesn't do a great job of identifying eGPU setups.

 

On what CPUs were you watching a video? Some intel CPUs came with a normal thermal paste between actual CPU die and the internal heat sink (IHS). Others intel CPU packagess have used a solder to secure the IHS to the CPU die. The solder does a better job of transferring heat from the die onto the IHS.

In regards to 3rd party vs. intel's own. I believe (and someone can correct me if I'm wrong) in regards to thermal rating, they're pretty much the same. However as you noticed there are sometimes imperfections (concavity for example) in the intel IHS that may not be present in a well manufactured third party IHS, which took care to machine things exactly as spec'ed.

These two variables play the biggest role in temp differential and help decide when to delid vs. not-delid and OEM IHS vs. third party IHS.

 

In regards to Liquid Metal (LM)... if I were to guess, since Silicon Lottery will seal the CPU IHS to the whole package, there is very little risk of the LM escaping and getting onto other electronic components inside the laptop.

However, with the use of LM between CPU package and an external heat sink/vapor chamber, sometimes the less pressured mountings of heat sink/vapor chamber, poor fit of that to the CPU package, improper protection against LM leaks, etc. there is more of a risk of a leak that may damage something when the laptop is sent to a customer (with all of the jostling and what not that occurs during shipment). Less likely for a catastrophic problem when the customer boots up if there is no LM on top of the CPU package. See Spartan's posts in the Area 51m R1 thread of what happens with a leaky LM application.

Again, I have no actual knowledge this is the case, but that reason alone seems to be the most logical.

 

Did you undervolt in this test? If so,. by how much?

Also. were you fans on full blast? performance, or balanced

 

This is actually a pretty good effort.

An NBC tester scored 6381 in a water cooled desktop with some crazy CL16 3200 RAM:

https://www.notebookcheck.net/Benchmarking-Intel-Comet-Lake-S-CPUs-with-up-to-10-cores.466287.0.html

 

When do we think we will see some actual reviews? I'm going to be pulling the trigger on one with the 10700k, 2070 Super and 300 Hz display but was hoping to get a better idea of the thermals with the vapor chamber cooling. Have seen a couple threads here of great temperatures but I don't know if that was stock, delidded, repasted with LM or what? I'm upgrading from a Lenovo Y740 with 9750H and RTX 2070 Max-Q with a beautiful 144 Hz 500 nit HDR display so I'm hoping this 300 Hz panel in the Area 51m R2 is still nice and bright.

Sent from my SM-G970W using Tapatalk

 

Probably not for at least a couple weeks my guess, and will be somewhat limited by what spec the system is. However the info on these forums is gold (thanks again @ratchetnclank) especially for those posting benchmarks. Also I believe @Spartan@HIDevolution said he was going to review his system once complete.

Unfortunately I do not have have the setup your interested in (4K, 2080S, 10900K), but will be certainly posting here my benchmarks and making comparisons to the last system (17 R5) I had after i pick up my system tomorrow!

I'll be sure to post benchmarks before and after any significant changes to the OS, whilst for the time being it will all be stock thermal solutions, except for a couple cheap low profile nvme heatsinks.

https://www.amazon.co.uk/dp/B07K9MF25F/ref=cm_sw_r_cp_apa_i_pAJiFb3TV8807

Only other additions im adding are an 860 EVO 1TB SSD and 970 EVO 1TB nvme.

 

I know this is off topic a bit but what are your guys opinion on an R3 with 3xxx GPUs? Will it happen or no? Its hard to drop another 4k on this R2 when the performance is minimal.

 

Thermals are unlikely to be of any concern with the 10700 and 2070.

 

A fair question, I think its almost certain to happen and the performance increase impressive, as you'd expect from a generational leap (as opposed to a mid leap with the Super series).

However it will be roughly a year after general release that the GPUs will be included in any R3 laptop form. Question is can you wait that long? Can you play Cyberpunk effectively on the system you already have?

Only you can answer that, overall we could always be waiting for a better spec system. My advice? Buy this current model to the best discount you can get, at what you can afford, keep the box and take good care of the system, then sell it to go towards the R3 if you feel the need.

 

Here's a legit Area-51m score by me:

 

Ampere laptops will be available 1 year after the release? That sounds way off.

RTX 2080 desktop launched on September 20, 2018, and I purchased my RTX MaxQ laptop in late March 2019. It was probably available with RTX even earlier. Area 51M with RTX was available in Jan 2019. A little more than 3 months after RTX 2080 launch.

I remember people who purchased the Pascal variant of the m15, which launched in Oct 2018, scratching their heads. Overall Dell was happy to launch three m15 generations in 2019. They are ruthless.

There are rumours that Ampere is already in production. It will likely launch around Sep, and we can reasonably expect to see the R3 in early 2021. Dell will push it out ASAP, they are almost surely working on it already. This is clearly the worst moment to purchase a new flagship in the GPU cycle. The GPU will be obsolete in a few months, with no upgrade path other than possibly through the AGA.

Edit: I literally wrote the above off the top, but now googled Ampere release date:

https://www.tomshardware.com/news/nvidia-rtx-3080-ampere-all-we-know

Sep 2020 release expected.

 

what settings? ie. Power management set to Adaptive or High Performance. Any changes to like the texture filtering quality or anything else or just default? That's 3000 more than what I get with my Area-51m with a 9700K/RTX 2080

 

Standard settings, nothing changed. Adaptive.

It's with 5ghz OC and slight OC on 2080.

Stock:

https://www.3dmark.com/fs/19316750

 

I stand corrected, thanks for the info, sorry I had it in my head the turnaround was far longer. Interesting...

 

The question is: will you be able to play Cyberpunk 2077 on the 51M R2 (which is effectively the same graphics-wise as the 51M R1)?

https://www.pcworld.com/article/356...ll-in-on-ray-tracing-and-nvidias-dlss-20.html

Very clear this will require Ampere at max settings, especially in a laptop where you can't brute force things by running your GPU at 300W or something like that.

 

-100mv undervolt. Fans at full blast.

 

It’s amusing that your DDR4 @ 2128 MHz sticks outperform R2’s blazing fast 2933MHz sticks in 3DM Firestrike Combined test