*OFFICIAL* Alienware 17 R4 Owner's Lounge

.

I use MX4 just fine with my 6820HK now running at -.03V and 4ghz with temps under 80C. You don't need exotic liquid metal paste to achieve a good result. The differences between one non conductive thermal paste and another are so insignificant that it's irrelevant.

 

Did you just try to assert little to no quantifiable difference between thermal paste products without supporting data while boasting ~80c temps?

 

SO first off, you're comparing a desktop processor, on a main board with around 2 times the PCB real-estate and resource cost for it's VRMs to provide clean, accurate power the CPU.
Second, the IHS has been a typical weak point for thermal transfer since Haswell. It's sole purpose is for faster OEM assembly by protecting the delicate silicon core. Devils Canyon (aka 4790K) specifically tried to address the 'overheating' problems of Intel's TIM and HIS interface. People resorted to de-lidding their processors (and still do) to work around this. I've no idea what you're on about that by not having a IHS that somehow makes BGA processors 'worse' but, you're wrong. They do nothing except add another layer that heat has to travel through to get exhausted by the HSF.
Third, you declocked your 6700K to 3.1ghz for testing purposes and, then either also under-volted it to below 1.0V or Intel's own automatic voltage control reigned it in(for reasons I'll get to in just a second) which, is well below the 1.1-1.2V of nominal voltage for a stock 6700HQ. Fun fact; less voltage means less power!
I suspect that you're also using a HSF designed to cool a 90W+ processor and not a 45W processor, which will definitely help let the processor run with less voltage given it's getting more cooling.

I don't think we can draw any conclusions from your testing. Other than you seem to enjoy emojis.

 

Yes, because it's been common knowledge for years. Do you need me to google that for you?

There are significant differences between liquid metal and non-conductive yes. In a sea of non-conductive pastes though, there is very little quantifiable difference between high quality pastes.

 

random question .. is the HDMI port in and out like the older models ?

 

The idea that "common knowledge" across "years" informs any sort of reality within tech is unreasonable. Actual common knowledge dictates that if you are going to make such a claim on a tech-oriented forum, you'd at provide data or at least an authoritative source if you want to be taken seriously. That aside, you may want to lay off of the "google" and pick up the books as you might learn something while giving your laptop a break from those ~80c temps.

 

1) IHS adds cost. Why should a lower powered BGA i7 cost the same as a more performing 6700K? And I have expected better temp without it. Would be nice for my setup.
2) Water cooling. Have you seen Clevo with it?
3) All 4 cores Skylake gen i7 should give about same performance vs. same clock speed.
4) Cooler master U3 can everybody buy. But don't expect the big gain in lowering your cpu temp.
5) Almost all people here undervolt their BGA processors whatsoever clock speed. Shouldn't they get the same gain in temp and power draw as my wimpy undervoltage? And why shouldn't I do the same for tests purposes?
6) Why should the cpu fan need to start up with a wimpy Cpu package power draw around 28-30W in tested bench if everything work as expected?
7) The numbers speak for themselves!!

 

Anyone with a 1070/4k screen:

How does the display handle games at lower resolutions? Is it as bad as standard definition on a HD display?

I have no experience with a 4k screen yet.

 

Mine hasn't arrived yet but considering 1080P divides into 4k exactly by a factor of 2 it should look pretty good since it doesn't have to do any interpolation. It can just pixel double.

 

Its factor 4 from 1080p to 2160p and it still interpolates

 

Depending on the type of upsampling this could make the screen very blury. Its crazy if this is the case. I hope at least every pixel from the virtual screen gets at least 1 pixel on the actual screen to mirror its color. It should get 4 pixels to just act like 1 pixel. That would be ideal. Its the opposite of how dsr works and thats why 4x dsr (2160 to 1080) looks so much better than any other ratio because each virtual pixel only weighs on the color of one screen pixel. Im gonna try putting my qhd on 720 and see if i get interpolation.

 

http://imgur.com/a/4AkBa

You know what though, you're right. NBR is a tech forum but, it's clearly not an overclocking or enthusiast forum where the intricacies of thermal paste deltas might be better known. Still, I think everyone can benefit form being more open minded. Myself included; I can be salty sometimes and, forgot to catch myself.

 

Quick questions before getting this laptop. If I plug my HDTV via hdmi and my 1920x1200 via mini display port to hdmi will this setup draw performance while gaming on one of the external screen?

I watched a dismantle video and read iunlock guide for repasting. It seems quite easy (I did repasting years ago but on desktops and am good with electronics: soldering Nintendo 3ds fuses for example). Would this be an easy task? With th price of these laptops I really don't want to screw up and be left with a faulty unit that I can't send back to Dell for repair..
Also, I am French and I am not really sure if I can get this stuff in France. Anyone living in France that would know where I should look for?

Envoyé de mon SM-N915FY en utilisant Tapatalk

 

Your statements are inaccurate from many angles.

First off Gelid and Kryonaut are paste and not liquid metal.
They are actually the best two pastes out there that are not metal.
When I say best I dont mean that in a matter of opinion I mean that in a matter of thermal conductivity and testing.
Go search tests done and you will see both Kryonaut and Gelid always have the lowest temps when it comes to paste.
Why is that? Simple, they have better thermal conductivity.
Again I encourage you to go look at some thermal paste round ups from multiple different sites. The numbers done lie.
I personally get highs 70s on a 4.0 overclock running OCCT with Kryonaut/Gelid.

Regarding "you don't need exotic liquid metal paste to achieve a good result."
That is technically true but to be fair and make sure proper information is released I will dive in to more details.
We have one system with Kryonaut (paste) which has a 4.0 overclock and it says around highs 70's/low 80s running OCCT for several hours.
We also have the same spec'd system running Conductonaut also at 4.0 on all cores staying in the mid 60s.
One only maxing at 63C.
So with that information a user can decide what a "good result" is.
I personally say going from 80s/90s to mid 60s as a "good result".

 

You are mis-informed and, you didn't even take the time to read my post. That's okay though because, I'll show you some evidence in just a moment. The delta is just 1 degree between kyronaught and MX4, across most high quality thermal paste's you're at most, looking at a 2-3 degree difference.
Case in point;
http://overclocking.guide/thermal-p...d-with-air-cooling-and-liquid-nitrogen-ln2/6/
1 degree between kyronaught and MX-4 which I might add, is in an almost completely 100% perfect scenario which removes all variables. Overclockguide used copper plates, a simulated (ie consistent) 160W thermal load and, PT100 thermometers(which cost several hundreds of dollars) . Unless you happen to have access to a professional laboratory, you're not going to have a more accurate test environment.
but just in case, let's also look at something more 'real world';
http://www.tomshardware.com/reviews/thermal-paste-performance-benchmark,3616-18.html
Again, just 1 degree difference between Gelid E. and, MX4.

THW particularly points out there's just 3.2 degrees separation from Gelid and, generic thermal paste.
Anecdotal evidence is meaningless frankly, particularly in a situation where application method actually matters. You're right though in that, number's don't lie.

 

After speaking to Dell earlier today it appears the monitor horizontal line issue is definitely being addressed, the rep mentioned a time frame of mid-Feb to March for possible replacements... Only time will tell.

 

Yep, this was also confirmed on Reddit here: https://www.reddit.com/r/Alienware/...4_owners_that_do_not_have_banding_on/dbi1ks3/

 

I didn't expect AW to move to the new chipset so fast; normally they have a few months lag. Of course, the one time I'm impressed with their timeliness is the one time the CPU upgrade literally doesn't matter...And since Optane won't be utilized in the upcoming models (as far as I aware), I'm hard-pressed to be excited for anything other than the heatsink fixes and new screen sourcing.

 

Has anyone actually seen a confirmation from Dell and can refer to that? So far I've only seen people claiming that Dell has confirmed this but no source. If not, it might be forgotten when the number guys start calculating the costs...

Sent from my ASUS_A001 using Tapatalk

 

I'm not that sensitive to ghosting ... but considering the response time of the IPS screen ... you will see ghosting at any refresh rate. Like I said, I don't notice it, either way. But I will say that the jump to 100hz refresh is night and day better than playing at 60hz ... the "perceived" increase in fluidity at the higher refresh rate makes a HUGE difference.

But be careful .. there are two 1080p IPS screens shipping with the 17's ... the LG one is the one that most users report being about to hit 90hz to 100hz ... mine happens to do 100hz fine.

Best of luck,
Joel

 

If any one is looking to buy an almost maxed out 17r4 I think I'm gonna sell mine .. is gtx 1080 i7 6820hk 512gb HD with 2 year warranty. Iunlock make it run amazing..Looking to get 2550


Sent from my iPhone using Tapatalk

 

It figures that when they announce the new procs is when my order goes through.

I do have a question though, how many external screens can the R4 drive?


My config on order:

Alienware 17 R4
6820HK / 32GB 2400MHz / 2x 950 Pro 256GB / 1TB 7200RPM / GTX 1070 / UHD/4k / Intel 8265NGW
Alienware Graphics Accelerator / EVGA GTX 1080 FTW

 

Anyone tested out the latest Bios update yet?

 

I wasnt saying it was not close, however in several independent tests the ones that were the lowest had similar spreads were Gelid and Kryo were always still lower temps then MX4)
To some people that extra degree or 3 is worth choosing the cooler different product.
Another thing to consider when choosing paste is the rate of degradation. All traditional pastes degrade and break down, it is in my experience the older pastes like AS5 and MX-4 degrade faster.
It is in my experience MX-4 while is great out of the gate, does indeed degrade faster then Gelid and Kryonaut.
I have had AS5 and MX-4 last 8 months before the temps were 10C+ higher then when the paste was applied.

Also this has nothing to do with being misinformed, I have personally used all of the products and I personally have seen better performance at high temps (i.e. decent overclocks) with Gelid and Kryonaut.
Now with all that said is MX-4 bad? No, it was the go to paste for a while, as was AS5. I was not intending to make the comment saying that it was a "bad" paste.
However for the few dollars more I would rather have paste that is a few degrees cooler and lasts a little longer.

 

Im using it


Sent from my iPhone using Tapatalk

 

@Lozz
http://forum.notebookreview.com/thr...3-owners-lounge.797458/page-205#post-10426704
Is 641 max as expected for @3.1GHz? Or is the normal more around 700cb like mine? I put up the bench numberes to show the difference.

 

Only out, no input so you can't use the laptop as a makeshift monitor.

200% scaling would give you the perfect scaling to 1080p

It's not bad, but your downsampled fhd game will look a little worse compared to a native fhd screen

performance will be the same

you can fly me to france and I'll do the repaste free for you

The sites are testing the thermal difference through a large desktop heatsink, or methods not suited for a laptop.

The Pascal alienware systems have an incredible amount of thermal density as the heatsink is very thin and small compared to, let's say, a clevo or msi GT heatsink. When a cooler is near it's thermal limits, liquid metal paste will help substantially more compared to a cooler that still has a lot of thermal headroom.

Try setting the display to 6 bit color in NVCP, these displays seem to not be true 8 bit color depth but 6+FRC, the FRC unit is what's causing issues with the line/banding.

This was pointed out by @D2 Ultima.

From the laptop you have mDP (DP 1.3) and HDMI (most likely HDMI 2.0), the thunderbolt can also output a signal directly or you use a hub to split the video signal. Not really sure on the USB 3.1 type C port.

From AGA it really depends on what card you put there.

I'm pretty sure you scored higher because of faster ram, stock DDR4 chips that come with laptop is absolute trash.

 

Pulled 160.310w in heroes of the storm with the 1070 using dsr.

 

You can't compare thermal paste results from the web done on test bench or desktop hardware and and say that these results are proper/correct for laptops with weaker cooling. This has been demonstrated several times by a lot of people.

 

FALSE. The "little quantifiable difference between high quality paste," part... because mx4 isn't quite high quality anymore to today's standard.

Not all thermal paste are created equal. To even pool them together as being similar is not correct.

Indeed.

LOL...ok hold on. There are several things you are missing here...

I'll just requote one of the most important factors that you are forgetting to take into consideration.

Did you read what @DeeX wrote? - "Another thing to consider when choosing paste is the rate of degradation. All traditional pastes degrade and break down, it is in my experience the older pastes like AS5 and MX-4 degrade faster."

With that being said, there are a lot of traditional pastes that will produce decent results when they are fresh. Have you seen my Crest Toothpaste challenge? It preformed great when freshly applied. So the up front tests only tell half the story.

No matter how perfect a test is, that has nothing to do with longevity and how a particular thermal paste will hold up with its integrity.

I'll show some evidence as well... and this is why a lot of us use liquid metal thermal paste to achieve the best temps.

11 Hour OCCT Test with Liquid Metal Thermal Paste:






There is a reason why I don't recommend mx4, as5 or any of the older pastes when there are much better and newer pastes out today like Kyronaut and Gelid (not the newest, but still among the top for LONGEVITY) with a proven track record.


I along with many of the enthusiasts on here are speaking from real world experience from long term tests.

I'd also advise that you or anyone that's trying to defend pastes like mx4 and as5, to not take it personally as the veteran's on here are only trying to help to provide you with better options.

It's important to understand that those older pastes are not even in the vocabulary among the enthusiasts.

As I always say...to each his own. If you're happy with mx4 then great. However, please refrain from providing inconclusive information on an outdated product that is inferior to what people could be using instead. That's the point.

@Papusan, @Mobius 1, @hmscott @zergslayer69 and many more who know this stuff to be true .... would all say the same.

"Not all thermal paste are created equal."

 

Before anyone brings up warranty.

1. The customer is always right
2. What they say can always be refuted by the customer
3. Look at the image below



















I personally run Kryonaut with a 190-200w draw Haswell E CPU, the thermal performance has been consistent from day one of application. For a laptop with much lower power draw, it will last much longer.

Liquid metal on the other hand, has virtually no dryout or pumpout that many regular paste suffer from.



And if "exotic thermal paste will not help temp" is your argument, please do try to beat the temps from my 12 hour OCCT run below on AUTO FANS and 22c ambient.

4.2 GHz will pull around 90w, similar to a 6700K at stock.

 

I really like to stress this point.

The Pascal alienware systems have an incredible amount of thermal density as the heatsink is very thin and small compared to, let's say, a clevo or msi GT heatsink. When a cooler is near it's thermal limits, liquid metal paste will help substantially more compared to a cooler that still has a lot of thermal headroom.



So if you take an Alienware 17 heatsink with full fanspeed and a 6820HK CPU clocked down to 22w (I have done this with turbo disabled), normal paste and LM paste will not make much difference.

But take the same CPU and clock it to around 90w (4.2GHz) and that's where regular paste will lose out big time compared to liquid metal.

On a lesser extent, buying better themal paste (ie: mx4 vs kryonaut) will net a bigger difference when used in laptops.

 

iunclock:

robert j prosser at gm a i l

no spaces

Thank you!

 

Hm, not sure. Mod MogRules from reddit seems to be confident, but who knows..

 

Exactly. +rep. I have tested several thermal paste's on my OC'd 4930mx. Only thermal paste that could handle my OC'd Hotwell Extreme is Liquid metal... Also the thermal difference between medium vs. good thermal paste is quite large if you are close to thermal limits on laptop hardware. I have been there and tested!! People see themselves blind on useless tests done on test bench with perfect fit on the metal - desktop hardware with a lot better cooling(these test results are more for desktop cooling).

 

Ok I'll challenge you, but I'll use mx4. No wait, maybe I'll use crest toothpaste. - Loser has to desolder off a 6700HQ and resolder on a 7820HK. Fair? Let's get them soldering tools ready...

Bingo...Bulls eyes. Well said.

But, medium grade thermal paste had lower temps on my 6700HQ delid. You're suppose to take a razer and cut the die off right? An even slice... then repaste?

 

Yes i can confirm that I just repasted with some old as5 figuring it would be better than the stock stamps and got worse performance. Opened it up 3 times to try different pasting methods too. Got some Kryonaut and thermal pads on the way.

 

as5 is conductive, but the bad kind of conductive

 

Our only "solution" is an Hdmi to usb3 capture dongle, which currently run about $80 or so. Unfortunate, seeing how alienware has the Hdmi-in patent.

Sent from my Nexus 6P using Tapatalk

 

how unfortunate really , my 2013 alienware worked perfectly with HDMI in and out with PS4
would you recommend me one of these capture dongle with that price ?
and do you they have any lag or delay when using these dongles on the laptop ?

 

Hello. I just received the 6820HK, GTX1070, FHD 1920 x 1080 Panel
Was looking for advice regarding the light bleed.
The pics are all at 100% brightness...

Suffice to say that the light bleed is really bad. Has anyone gotten a replacement due to this issue?

 

Btw, does anyone know if AW17 r4 has a 'battery life extension mode', when the battery is not charged to 100%, but only up to 50-80%? Didn't find anything related neither in BIOS nor in AW software. I understand that this laptop is mainly used from AC, but still it would be nice to prolong battery's life.

Edit: I believe I've found the solution, this mode should be available if OSD driver is installed. So this dialog will be available (from 15 sec):

 

So, I've now tried to OC and put my 6820HK up to level 3 and played Crysis 2 for 1-2 hours. Had HWInfo up at the same time which noticed that several of the cores reached 99 degrees and thus started throttling. Average temperature was around 80 degrees though. Did not change the fan profile from default value in BIOS. Is this so bad I should just return it? This is with a GTX 1080.

 

i haven't used one personally, but I've been shopping for one and the reviews usually do indicate about 1/8 second input lag on the more affordable ones, which may only be acceptable on non competitive or non fps games... so I wait.



The Magewell XI100DUSB was the one the looked like it was lag free but is almost $300.


Sent from my Nexus 6P using Tapatalk

 

Okay so I'd like to update you all with what a Dell thermal fix is capable of for AW 17 R4 GTX 1080.

Since I'm still waiting for Mobius to take my laptop for repaste, I thought I'll escalate to Azor Frank my abysmal thermal performance. (90 C when zero overclock. 99 C when OC level 3, 10 C difference across cores). Today, a technician came over and replaced the thermal sink/thermal pads and repasted my CPU/GPU.

On stock levels (no OC), temps are 70 C max. With OC level 1 (no undervolting) temps are 78 C. With OC level 1 (-0.1 undervolt) temps are 72 C. < 5 C difference across cores. Now I know this is nowhere near the performance team LHz has achieved, but I do appreciate Dell trying to fix things to atleast a resonable level, a huge improvement over the abysmal state the laptops were in before!

 

Hi. I received my Alienware 17R4 1080 last week. I have not overclocked it yet. I feel stupid but I am trying to send a pm to iunlock regarding repasting but I don't see where I can do that. Can anyone assist? Thanks

 

Has anyone experienced any Tobi IR tracking issues with their Alienware laptop? I have the 17 R4 and the software does not load when I try to start it. It does not stay in the system tray. I had Tobi support currently looking into it (we both did the same solution steps, uninstall the software and camera driver, remove any files, registry entries and resinstalled) but the issue is still occuring. Tobi support gathered log files and we will be having another remote session with the developers next week. I was just wonderng if anyone else had this problem and were they able to resolve it. Thanks.

 

Click on his username, and then click "Start a conversation".