*OFFICIAL* Alienware 17 R4 Owner's Lounge

Unbelievable, returned the AW15 R4 due to fault screen, returned the AW 17 R4 for a replacement due to more than 20 degrees core temperature diffs and just got the new AW 17 R4....and again notice more than 20 degrees diff between cores and CPU throttling when temps hit 95+ degrees. Should I just give up and switch brand?

And no, I refuse to repaste a machine costing this much just because Dell can't do their job building it. So pissed right now

Edit:

Some more info. On both AW 17 R4 it is core 1 and 3 that is more than 20 degrees above the rest. This can't be a coincidence. I'll take a wild guess here, those are the 2 cores mostly affected by a badly mounted CPU cooling?

Also noticed that my replacement machine seemed to have a service tag lower than the one it replaces (starting with a 3 instead of a 4), are these codes random or is it an indication that Dell infact sent me a refurbished machine with the same problem as then one I returned?

BTW, both the new and the old machine reports CPU fan failure when running the BIOS Diagnostics.

 

It's bs. Read my post on page 341. They are sending out defective machines and all beacause some idiot thought it was a good idea for the CPU and gpu to share heat pipes...

 

Several OEM's use the unwanted shared heatsink/pipes design in their laptop models. But AW take it longer and designed their new Not so smart Tripod cpu heatsink mounting on top of the unwanted design Put it the other way... A design flaw.

 

I think the cooling system is good .. just bad paste job my temps seem amazing after iunlock re paste



Sent from my iPhone using Tapatalk

 

Don't compare Dell's thermal paste and the job done by Dellienware fabrics or the hired outsourced tech team with what you can do yourself With tweaking heatsink and liquid metal. You need to fix almost every damn laptop pushed out from Dell's factory. And 36x on processor who is a lot lower than Dell's advertised OC profil doesn't push the processor at all. Put full load on both cpu and graphics with max OC 24/7, lets say +42-43x and you will see the temp will increase

 

I'm not trying to compare I'm just saying the system can cool very well if done right


Sent from my iPhone using Tapatalk

 

I know that, I only talked about this mess should be done the other way. Aka re-design the Tripod to a Fourpod cpu heatsink like other OEM's have and same time use proper pads and thermal paste. That's Dellienware now push out now isn't good enough. I'm sure this big mess wouldn't be so massive with better heatsink design.

 

Hey now...you've seen plenty of OC tests via OCCT so I think you're being a little too nit picky there. 36x on all four cores is a ball park of what we consider stock and a good daily driver clock for regular stuff. You don't need to have high clocks to check email etc... so your request is a bit irrelevant and obviously nit picking.

What's the use of OC'ing both the CPU and GPU for 24 hours? That is a bit unrealistic and you know that. Temps will increase? Likely a bit with that extreme test, but again why?

Guess what? Your temps will rise too...

Maybe I need to refresh your memory?



Maybe you have forgotten about this too...11 Hour OCCT Test...



95W at 4.3 for wPrime...



What are your temps?

 

If you mean the new heatsink design from Dell aka the well known Tripod is top notch, Oh 'well. Then I expext most part of the laptops pushed out from Dell's fabrics will have no problems with the cooling for Dell's advertised max OC profile!!

I don't run stock 4.0GHz - 4.3GHz clockspeed. My lowest dayly clock speed in TS is 4.5GHz.
From @4.8GHz bench in Wprime 1024M and about +110W CPU package power... + a nice score on top

 

Top notch? You're joking right? Where have you been hiding? I've been the one pointing out the design flaw, so I don't get your assumption.

Always trying to cause trouble lol...

There you go again. 95w on a mobile chip is more impressive then your 110w.

You know the difference between a mobile chip and a desktop right? Maybe not..

Anyway, when I get my 7700K let's go race.

::iunlock::

 

You should read my post before making comments bro I just pointed out the same as you say now. Why make a major case of that Dell screwed up and created a awful HS design for CPU? The heatsink design is screwed and shouldn't be used!! And if people neeed Dell's hired outsourced tech team to fix their machines.... You will NEVER get a good enough result. Would be much easier if DEllienware used a more normal HS design aka Fourpod for cpu. And not the flawed Tripod design

Edit: Everybody should know for long time ago that the useless Dell's hired outsourced tech team or depot will screw up everything they touch. With the Tripod design... Guaranteed!!

 

That's my point and the design flaw has already been pointed out. Nothing new. I have many post about this.

You're opening sentence makes no sense. I had replied, because I did read it lol...

Lost in translation I guess..

::iunlock::

 

This ↓↓↓ is my first reply. And why make a BIG case of this? I know I'm not the best in English , You know that. Because you have said My English bullied isn't good several times before But I mean this below is easy to understand

 

Not sure if I have an issue or not.

During heavy gaming my cpu core temps max out at 65 degrees with with a differential of 3-4 degrees. My gpu temps never go above 65 degrees. The problem is my pch temps which are always 65-72 degrees. Are the pch temps something to be concerned with?

 

Not with this temperature

 

I played Watchdogs 2 for maybe 30 mins on my r4 i7-6820hk and gtx 1080, played at qHD resolution, everything maxed out, laptop plugged into power supply. Max temp on Core 0 is 98, Core 1 max 92, Core 2 max 93, Core 3 max 89, Package at max 98. The graphics card max was 85 degrees. Processor Max clock was only at 3595 MHz. My build is from late December, are these temps and clock speeds normal? Dell says not to worry about the temps at all, that the processor can reach excess of 100 degrees and the video card can be over 90 degrees. I worry since mine is nearly at 100 degrees on the processor and just about 90 degrees on the video card. The system is brand new...worried I got a lemon.

 

Not good. And as expected. Run Dell's advertised maximum overclocking profile for the processor. See how it goes @iunlock

Edit: http://forum.notebookreview.com/thr...ven-core-temps-due-to-uneven-heatsink.797477/

 

Do I access that through the Bios? Sorry. Completely new at this...thank you for your patience!!!

 

I'm not too familiar on the new BGA machines from Dell. But take a look http://forum.notebookreview.com/thr...-settings-for-4-0ghz-to-4-3ghz-stable.787791/ in post #1 or ask the other.

 

I'm not making a big case...just miscommunication. No worries. Of course we know the tripod design was a big mistake. Such a shame. - And I'm not attacking your English, trust me that's the last thing I would do as I speak several languages myself. Heck, your English is better than most Merikans.

Stock out of the box with power limits unlocked?...Pooooof....haha. Dell Exclusive.

Stock out of the box with stock clocks and staying under 45w...mmmm...it'll be fine.

 

Thanks Or @iunlock shall we say 감사

For the second. This ↓↓↓ is

Spoiler: Stock out of the box with power limits unlocked

My first experience with the new AW BGA models. Now we are in the beginning of February and similar continues Mr. Azor need a kick in the butt.

 

It's been a hit or a miss with the new systems. The past 5 that I worked on last week were all from the new batch. 3 out of 5 of them were mmmm okay in temps and not too bad for running stock. The other 2 were just crap. So I think they are coming from different assembly lines and some are good while some are not.

AW/Dell better address this soon because the new slogan that will be branded on them will be, "Gaming with High Compromise. You can count on it."

 

You say running stock. Maximum OC profile? If not are the results even worse.
We who have had the older socket AW machines before, have never experienced that we have to use other to fix Dell's flaws. This is something new.

 

Running at stock clocks with power limits lifted. 3 of them were okay temps and with not too bad of core differentials. 2 of the machines were just rubbish like the original batch.....Quite sad....

 

And what happens when you run all of them with Dell's advertised max OC profile?(think that's 1 [email protected] and 4 [email protected])

 

Oh I never use the stock OC profile presets. Honestly, I think I've tried them maybe once on the previous gen AW lol....as you know those OC profiles are absolutely a joke.

I always start at 36x across all four cores and for wPrime the power limit is always unlocked before running that. Then my next increment is to 40x, then 41x, 42x and 43x if it can hold. The 3 laptops were in the mid to high 80C's while the other 2 were all in the 100C mile club LOL... sure enough when I had opened them up, either the thermal pads were not properly placed and/or the HS was just way off so it had to be adjusted.

After the fix they all ran fine, but still... shame.

 

That's not good at all. After more time running at those temps the system will eventually thermal throttle down to 0.76 GHz. You can either follow the guide to re-paste it yourself and see if you have any luck or Call Dell and see if they can send someone out or send it to them both may or may not fix the issue, mine didn't. I have seen pics of iunlocks work and it seems as if he knows how to fix it but Dell is sending the bulk of the systems out broken and most of the time can't fix them in my experience. I would call Dell if you don't repaste yourself to get the repair process running before you return period runs out.

 

Are all 17r4 made in the same facility? My shipping info says Tennessee, saw another post reference California. Perhaps there are correlations to explore.

 

Yea it's a toss up. If someone isn't experienced then they risk messing it up even more and/or one can almost guarantee that a Dell Monkey will surely mess it up in one way or another. I've encouraged a lot of my international clients to do an exchange until they get one that is somewhat kinda sorta decent. At least that way they have a better base to work from knowing that the heat sink isn't as warped or messed up as the previous unit that had bigger core differentials.

I've also spent hours on TeamViewer with my clients overseas in Europe and Australia, trying to coach them step by step, but that doesn't always work either. i have one client in Europe that threw in the towel and sent his unit back and just ordered one with the help of a relative in the States that will take it back with them. Kinda works out well if all the cards fall into place, but yea it just sucks that there's this much QC control issues.

To be fair, there are some newer units that are halfway decent so at least there is that going...

 

Not sure but I'm getting these events too.

Man I was just starting to like these things and all of a sudden the bios version that gives me the best performance turns out to be a ticking time bomb that also drains the battery while I play.

I'll have to do some testing later to try to get to the bottom of things. Been up all night.

 

So, is it safe to assume that if core 1 and 3 is way hotter than 2 and 4, it is due to a faulty mounted heatsink? Have seen this on 2 AW17s in a row now, can't be a coincidence.

 

Wouldn't "you CAN'T count on it" be more appropriate? That said it would probably work as a pocket calculator at least

 

More likely the CPU throttling than anything to do with the GPU

Sent from my SM-G935F using Tapatalk

 

That would be one heavy calculator haha.

The slogan means: ""Gaming with High Compromise. You can count on it (that it'll be with high compromise.)."

It's a sleazy marketing move to confuse people. Can't give it away too easily eh?

 

Dellienware's previous slogan... Fully *Throttle* gaming Maybe its still valid Throttle or throttling can be explained differently

 

Nope, 90% of their laptops are sent out needing work.

 

Not sure if you understood my question, it was if core 1 and 3 was the most likely to be affected by the uneven heat sink. Have seen other screen shoots that also seem to have core 1 and 3 run hotter.

 

It's very possible that it could be an uneven heat sink or just bad thermal pasting, or both. Alienware is not opposed to you repeating it yourself as long as you know what your doing or have it fixed through them. Sadly even if they address that issues and the cores are even it may still suffer from the wide spread thermal issues. It's very hit and miss. I wish you the best of luck.

 

Not sure but I believe that you might still be missing the point here. As I assume that the CPU is mounted the same everytime, it would seem reasonable that the same CPU cores were affected each time if the heat sink was badly mounted. On both AW17 R4 I've tried it has been Core 1 and 3, and on screen shoots posted by others, 1 and 3 seem to be the ones mostly affected there also. A bad pasting or similar should affect all cores as much or at least be more random in how each core was affected.

 

I noticed something interesting, that affected both my AW17R4s. If running the BIOS Diagnostics without OC disabled, it would pass the CPU fan test, but if OC is enabled, and one of the 3 profiles are selected (or probably the custom one as well), it would indicated CPU fan failure. This I could repeat on both my (soon to be returned) AW17s, do you guys see the same problem? THis was with both bios 1.08 and 1.09

 

Its a combination of bad paste, uneven heatsink and the fact the cores 1/3 take the most load. I had like a 25c difference between mine. After a repaste, i dropped all 4 cores a good 20c. I still hade a 15c core differential though. Bending the top tension arm to put more pressure on the top of the heatsink managed to knock about 5c off of core 1/3, but had no effect on the bottom two. It is random how each person is effected, some people dont have crazy differences on theres. Instead they just have all 4 cores running at 90+. Which would point more towards a ****ty paste job. I personally just had a massive core differential because of a messed up heatsink. My 1/3 was in the 100+ range while 0/2 were in the "acceptable" 80-85c.

 

Haswell was full throttle gaming for sure. 800mhz at 76°c. And now you can game without compromise as you call tech support and consider a refund.

 

Assuming that you mainly use an AW17 R4 for gaming, when would unstable overclocking of the 6820HK CPU be a real problem if it is equipped with a GTX1080? Ie when would the lack of that extra CPU performance that could have been squeezed out by OC:ing start to become the limiting factor?

 

Do you guys know why intel XTU does not save undervolting after sleep mode or reboot even if it says it is applied? Any solution?

 

Honestly for just straight gaming you really dont need a high overclock. Not many games actually take full advantage out of your cpu. Ive personally noticed with mine 3.8/3.9 seems to be the sweet spot. I can get about a 10-15fps boost at those clocks from games that are more cpu heavy like say modded skyrim/crysis 3. Once i start clocking it higher than 3.9 i really dont see much of an improvement besides a few 1-2 frames here and there. To me the higher temps arent worth the slight improvement. The only way i see the 6820hk would start becoming a real bottleneck at lower clocks is in the future when you use something like the graphics amplifier with a higher end card. But even then you would take the GPU inside the laptop out of the mix and you would see lower temps overall to begin with, which would in theory allow you to clock it higher.

 

Hey guys, I'm fairly new to this forum. I just got my first AW ever - the 17 R4 (7820HK, Nvidia 1070) about 2 weeks ago and just had some time to play with it. Been reading parts of this thread and been trying to make sure I don't have a lemon here. At stock speed, prime95 small FFT, I am getting temp of the 4 cores between 59 and 68 (usually more like 60-65 degrees), with the average difference between the cores to be about 4-5 degrees C. Is this normal? Core 0 and 2 seem to always be the highest and core 1 and 3 seem to always be the lowest, towards the lower end of that range. I ran this test over 15 minutes and just basically sat and watched my temps change.

Edit: I should mention that as soon as I stop the test, temps range from 38-40, with average difference really being about 1 degree C.

I set it to level 3 OC - the default settings and I am getting Core 0 and 2 averaging mid 80s to 92 degrees and core 1 and 3 averaging from 76 to 79 degrees C. These temps seem really high and seems like a therma paste problem?

 

It varries from game to game. Turn on gpu-z and have it plot your gpu useage durring a game. If the gpu doesnt recieve commands from the game's cpu code fast enough it will be sitting below 90% useage.

There are many exceptions though. Overwatch uses a multithreaded renderer so one thread computes the core game processing and another thread ques up and sends commands to the gpu as it becomes ready. This way the core code doesnt wait for the gpu to accept the next command and the gpu doesnt wait as much for the core code to send out the next command when it is ready. When games start using vulkan and directX12 fully they will do something similar but the gpu will handle the same job as the cpu based dedicated renderer thread does now.

But for the most part if the gpu is in the 90-100% range it is keeping up with the cpu and below that is where you might want to increase your graphics settings or overclock your cpu.

 

Perhaps a better question than the one I entered above would be, if the OCing on the AW17R4 is crap because of badly mounted heat sink, paste or any other of the numorous problems plaguing this machine, how much of a problem is it really when it comes to gaming compability?

Of course, I might just return it anyways just because I don't wont Dell to have my money for producing something with this crap quality, but then I need to get something new and their really isn't all that much perfect gaming laptops out there. If MSIs designers hadn't been blind, that might have been the solution, but they just look like something a teenager would create after having been stuck in a meth lab for a week...

 

For the games that i play i get anywhere between 144hz and 100hz with my gpu set near the cpu bottleneck and my 6700hq is locked at 3.1 ghz so you should be ok depending on your standards but i mean it just would suck for me to know that my machine could do better (and im sure it could if the 6700hq was unlocked) if just some minor changes were made in the assembly.

 

I agree. But with the average AW17 R4, it can't due to crap production