Alienware 17 R4 GTX1080 in game random Stutter & Freeze

that's alot of misinformation you are putting out, VRMs can withstand over 100C no problem. when you get a bad VRM no amount of thermal pad will save you. That is why my old 1080 will stutter under high wattage regardless of VRM temp, it is just a faulty VRM and needs to be replaced.

 

Not so fast my friend. And yoo as well expect the components used on laptop hardware is of same or better quality + the cooling is equal as desktop hardware? Oh'well

 

Care to explain why the same EVGA company u featured issued thermal pads to GTX1080 desktop cards due to their cards dying from VRM overheating failure?

https://www.evga.com/thermalmod/
https://www.kitguru.net/components/...o-issue-diy-fix-for-gtx-1080-vrm-heat-issues/

again, this is official evga vs your imagination. your old 1080 stuttered hopelessly probably because you removed all the thermal pads from day 1 thinking those thermal pads are placed there for show and you dont need them. find one GTX1080 manufacturer that makes GTX1080 without thermal pads on GDDR5x, VRM and chokes then we are convinced, other than your imaginary laptop of course.

 

my old 1080 stuttered hopelessly with fujipoly 17 pads on all VRM DDR5x chocks and they make full contacts with the heatsink so yeah it is definitely my imagination. Maybe if I imagine hard enough it will stop stuttering like yours. Or I could ask dell and get a new motherboard without faulty VRM

 

Easy bois, no need to get worked up.

To back @InvoluntorySoul, it does seem that my previous unit, no matter what I did cooling-wise (and I did everything at least twice) it WOULD NOT work. The freezes were always coming back pretty much at the same time and with the same intensity. They had this same intensity when the notebook was brand new from Alienware and I first played Doom and the CPU reached 102ºC and kept the same sh*t after heatsink lapping, bendind, Arctic pads, proper contact, liquid metal and PCH mod.
Some units are just more sensitive. My new one has stock pads and everything works fine, temps are much better, you can feel the difference on the keyboard easily..

And, in defense of @alexnvidia, you're right. You shouldn't use the PC without any thermal pads but I have no doubt it would work. It's just a question of for how much time.

About that EVGA stuff. They were both in the wrong.
The VRM manufacturer sent them lemons, no doubt. But eVGA, with their awesome heatsink design, were also not properly conditioning the VRMs. Most other manufacturers have heatsinks that fully cover the VRMs.
Bad VRMs + Bad cooling = boom

 

Wait, so are you guys saying this is a the faulty mainboards fault, or gtx 1080's?

Think I should pester Dell support for a 1080 exchange, till I get a good one?

 

If you continue to use your laptop when it's stuttering due to power delivery components overheating, you might end up with permanent damage which will require a motherboard replacement. The steps detailed in this thread is not for the average person, so if you feel that you are not up for it and prefer to deal with dell outsourced India moron tech support, by all means pester dell. But if you are up for it, I totally fixed my stuttering issue without any component replacements other than thermal pads.

 

evga card blow up because the mosfet were faulty, not because they were uncooled

 

So I got my new AW 17 R4 and overall what an amazing machine. However, mine has the stuttering issue. It reared its head while running OCCT and Heaven for 10 to 15 minutes. Cooled laptop and let it come back up and it started stuttering again after roughly the same amount of time. I made a post on the Alienware reddit https://www.reddit.com/r/Alienware/comments/6gso6n/brand_new_17r4_stuttering/ and am not sure what to do.

I have until 6/29 to return the machine and am really thinking I should do that and just get my money back without going into the hassle of tech support and trying to repad which might not even fix the issue. Anyone have any advice?

 

That's exactly what happened to my laptop as well before I repad it and fixed it completely. You can always try your luck and let dell handle it or you can take matters into your own hands if you are up for it. Regardless of your choice, if you still want an Alienware, you are probably gonna have to repaste it at some point so it's definitely worth a try. And you shouldn't stress your system in this state for extended periods of time to avoid permanent damage

 

Surely someone has a laptop with solid performance? Or perhaps no one is testing to this extreme? BF1 causes the issue though so you would think that it is being uncovered regularly. You don't think that there are good samples with solid mosfets and pads in them?

Also, thanks for the tip but I talk about not continuing the stress in the reddit post.

 

If you really fancy the 17 R4 I would return it and get a new unit. Repeat this until you get a unit that DOES NOT stutter out of the box.
Repading and repaste is reccommended. But doing so in a unit that doesn't already stutter from the factory will yield much better results and you will be more certainly problem free in the long run.
Alex I see what you say but I think you were just lucky. I wouldn't put my hand in the fire betting your unit will never have stuttering again. It has become obvious that some units are more sensible than others. Might as well return, use and abuse Dells Premium Service until you get a fully working unit. When and if that day comes do a repad and repaste and you'll never have to worry about this sh*t anymore.
Plus, new units have better heatsinks. They are discontinuing the problematic CCI ones and replacing them with Sunon units.

Your laptop should run at 100% from the factory. You wouldn't accept a badly running car just because you can, probably, fix it yourself.

Prestone, if you can't be arsed just ask for your money back. The 17 R4 is a wonderfully built machine when it's been properly quality controlled. But there's plenty of healthy competition out there. Check notebookcheck's top 10 gaming notebooks for good alternatives.

 

I can tell you my experience. My first unit stuttered.
I tried everything, been here since the beginning. Almost two solid months of troubleshooting.
Repadded, repasted, tried various pastes, liquid metal, I lapped and bent my heatsink, even used thermal paste in some of the power delivery circuits to ensure best contact.
PCH mod, cooling base, HOURS of software troubleshooting, BIOS upgrades and downgrades, vBIOS too.
Nothing I did worked. the issue ALWAYS came back.

After two motherboard replacements, one new heatsink and one new SSD I finally have a 100% working unit with none of the hassle.
It's on stock pads and it works freakin' fine.
Mobius1 even has one with stock pads and paste running at 5Ghz, albeit dangerously hot! But no issues.
Some just came with lemons for power delivery components, just trade it until you get a problem free unit, they do exist.

 

I'm kinda with you on this.

Also, once I'm out of the return window I'm pretty stuck. Hours of Dell "techs" "fixing" things and replacing motherboards and heat sinks. I guess I can buy and return/refund the laptop until I get a good one right?

This sucks lol. When the AW 17R4 is running right it's far and away the best laptop on the market. Why would they cheap out on some 50 cent mosfets and cooling pads.

s
m
dang
h

 

Install this BIOS 1.0.5 http://www.filedropper.com/alienware15r317r4s1012k105

Install this vBIOS http://www.filedropper.com/vbiosdowngrade17r41080

Stuttering should be gone after doing both of those. That's what fixed it for me.

 

Apparently that BIOS is 30W less than the current BIOS. I'm not sure that turning it down is actually a fix.

 

Maybe this little analogy will help.

"Hey Ferrari, my steering wheel vibrates like crazy at 240mph and the engine goes to neutral, t's not safe!"

Everyone in NBR:
"Don't worry dear customer, just install this engine management downgrade from us. Don't even stress if the new ones fix existing issues or have better fan curves .
And install this gearbox management downgrade we got from God knows where. We take no responsibility if you brick your car....
This way your car will never go faster than 220mph, hey, it's fast enough right?! Right?!
Who cares if the cooling unit is broken and your pistons are made from wood, just downgrade the hell out of it, should last until the warranty is over no problem!"

 

the stutter problem is very hard for dell to test, someone actually has to sit behind the computer playing a game to notice it. Dell prob just got really unlucky that a batch of their i7 7820 / GTX 1080 motherboards have faulty VRM or power delivery unit. My newer replacement (made in 5/2017) has 0 stutter with OCvbio, with OCvbio this thing can hold over 2000Mhz core constantly where under old vbio it will only hold 1850Mhz.

When you play game run GPUZ and look for PerfCap reason and GPU load. If when the reason is 4 and GPU load would drop to 0% and GPU power also drop to like 120-130W it means your VRM cannot handle high voltage and you are better off replace the VRM. Thermal pads are not going to magically save a faulty VRM. You notice only 2 people so far claim they are able to fix stuttering without replacement I would take their words with a grain of salt

You don't have to deal with dell support, just return for refund and order a new one. Don't bother with repaste unless it is stutter free in OCvbio stock.

 

Turning down the wattage lowers the temps and prevents the overheating, while only performing 2% worse than with the OC BIOS. Seems like a good trade off.

 

Yeah I think this is where I'm going. I put a tweet out to AW Umar to see if he can help or give a suggestion.

Dell should be able to get this to pop up near instantly in testing. Load GPU and CPU up at the same time and wait 10 minutes. I ran OCCT and Heaven. They could do something similar. I think you're right though that bad VRMs exist and that's probably the main cause. Even if it is a pad contact issue, I don't want to miss the return window trying to figure that out.

 

You know what beats that trade-off? A trade- in, for a new unit, one without issues.

They all ship with the OC vBIOS, there is no point for customers to install older vBIOS dumped from someone's computer.

 

I hardly think this has anything to do with luck. @Shark00n , your unit probably suffered permenant damage from continuous use or testing after the stuttering first happened. When my stuttering first started, i stopped playing less than 10min immediately and i repeated the test at most 3 times before concluding i have to do something about it. even after the first attempt sealing the gaps with thicker thermal pads, i still experienced stuttering though the time for it to start happening has prolonged. Took me 3 attempts to completely fixed it. not 1 but 3. even the most experienced ones can't get it right the first time.

it's been way over a month since i started this thread on how to fix the stuttering issue and my laptop has remained stutter free from all the games i played, so luck has absolutely nothing to do with it. As you can all see from the video interview i did with ownordisown, i captured a video on BF1 showing stuttering happening all over the place. Stephen actually cut the video short otherwise you can see a lot more stuttering. If this is a batch problem, it doesn't explain why some owners who got their laptop as recent as this week still suffered the same stuttering issue. and mind you all, my GTX 1080 is full time overclocked 180MHz on GPU and 500MHz on GDDR5x on top of the already factory OC vBIOS. my GPU power consumption regularly peaked over 200W during gaming sessions. this is not a miracle, not a batch problem, but pure hard work and dedication. Some OCD characteristic helps a lot too.

The VRM and GDDR5x needs to be cooled properly despite some imaginary believe saying they dont need to. No sane engineer in this world would spend time to design cooling specifically for those components if they dont need cooling. In fact, if they can get away with it, they would definately do that (example: PCH).

Identifying the problematic area is not something easy and from my experience, it's actually very time consuming and difficult. you need to be really patient, and there are factors like resuing thermal pads which can cause problems too. this was what i found out later on and caught it by the "wiggle test". I agree with @Shark00n , flashing the old vBIOS is not the solution. This laptop is designed to go as far as 180W.

For those who are not up for it, just let dell handle this mess. I believe in if you want things done right, do it yourself. but that's just me.

 

I'm not too sure man. I literally repasted for the first time just 5 mins after the stuttering started. I was checking for proper contact every time.
I changed thermal pads soon after.
Plus, my supposedly damaged unit worked fine with the old vBIOS so I really don't think I damaged any parts. It came borked from the factory.

 

i just love how some imaginary think they can fix faulty components with some thermal pads. The fact is you have 10% chance to fix the stutter with thermal pads or 100% chance to fix with replacement

 

Some new weird information.

When the laptop is sitting flat on a flat surface the temps of the CPU are between 85 and 92 and the clocks are 2.7 to 2.8. The GPU is up around 78 or 80. At these temps/clocks there is no stutter in Heaven bench with OCCT running.

When the laptop is propped up about 1/2 inch in the back it gets better airflow and lower temps and runs the clock up a bit. These temps are 78 to 85 C and the GPU is around 70 C. The clocks that the computer sets are around 3.0 to 3.02. Under these conditions with cooler overall ICs the laptop stutters in Heaven bench.



This is really weird. Only thing I can go back to is overloading the voltage delivery components. :/

Not sure where to go from here.

 

Maybe you can try removing all the thermal pads and let your laptop fly stutter free to dreamland. One misleading member here vouch for his unicorn laptop that can work without thermal pads, perhaps even without heatsink entirely! He should publish his work to the engineering community, imagine how much money every gpu maker company can save omitting thermal pads and heatsink in their design. Wait before you proceed, this is a joke. Don't do something like that. It's just plain stupid.

Joking aside, you are on to something. Have you tried checking the gpu power consumption when you lifted the back of your laptop? I'm thinking the reason could be when you lifted the laptop, your gpu temp is lower, allowing it to increase power consumption for not hitting thermal throttling limit. Thus the VRM needs to work harder to provide more power, making them hotter without proper cooling. Just a hunch.

Without actual airflow underneath, lifting the back doesn't necessarily means cooler components especially for power delivery components like VRM which are covered underneath the heatsink and thermal pads. Also it's worth checking out the gddr5x thermal pad contacts too.

Again, if you are not comfortable doing this, let dell handle this mess. This procedure, is much more difficult than just doing a simple repaste and most people will not get it right the first time and it can be very frustrating. I learn this the hard way. Took me 3 attempts to get it right, and each time before that I was darn convinced I did everything right only to find out other gaps the next time I disassembled it. But once you get it right, you are rewarded with a peace of mind and a truly beast of a laptop.

 

man do i ever feel lucky. My new replacement runs pretty dang good after a repaste + cooling pad. 4.3ghz is stable for games with temps maxing at 85C. There are some core differentials that can vary up to 9-10C, but i'm not going to stress about it. I also have the suncon heatsink.

Bottom line is that the horizontal banding is gone, the cpu does not bounce off thermal limits and none of this dreaded stutter i've been reading about. Thanks Dell! Great customer service

 

Happy for you! Though I wasn't expecting sunon heatsink to have cpu core temperature differential problem.

 

smart move getting the replacement and not trying to fix yourself unlike one misleading member here constantly vouching for his unicorn thermal pad replacement that can fix faulty components

 

Get a room you two

 

Hi. Wondering what motherboard you have? Did they replace it? Mine lists BAP20 LA-D753P Rev 1.0 (A00) 2016-10-20. Wondering if this board has bad VRM components.

 

Man my motherboard is Rev 01. I wonder if there's bad parts on Rev 01 motherboards.

Do you have a pic or the part numbers/build date of the motherboard?

 

Hey guys, signed up to share my experience

Got my Alienware 17 R4 Thursday of last week from Amazon.com
I have the 7820HK, 32GB RAM, 256 SSD+1TB HDD version.

Pretty much the only game I've been playing the Witcher 3. I noticed the stuttering issue a couple days ago. I haven't modified any of the CPU/GPU clocks or voltages. I've also just updated my BIOS using the version released from dell on June 14th.

The exact way the issue reproduces for me: after about 5-10 minutes of gameplay, the screen will freeze for 1-3 seconds while audio continues playing. This temporary freeze happens every 20 seconds or so. Even if I alt-tab back to the Windows desktop and leave the game running in the background, the screen will still freeze for 1-3 seconds every 20 seconds, so it doesn't seem to matter whether the game is in the foreground or not.

I should note, that the stuttering only started after I downloaded HWMonitor to inspect my temps. I saw that my CPU was setting a max of about 90c-97c, and the GPU was maxing about 80c. I considered this to be too hot, esp for the CPU, so I started to prop up the back of the laptop on whatever items I had nearby (usually a remote for a TV). This improved the temperatures to about 80c max for the CPU and 69c max for the GPU. But trying to lower the temperatures in this way seems to introduce the stuttering issue. When I got back to letting the laptop lay flat on the surface, the stuttering goes away....

Is there anyone else who might be propping up the back of their laptop and seeing this issue only when its propped up?

I would consider letting it lay flat to be a solution I am OK with, if only the CPU temps weren't pushing into the 90s.

 

Maybe someone can correct me on this, but there is no known other revision of motherboard so far. they are all the same A00 with the same date. Though i strongly don't think this is a "bad VRM" issue, it's best for you to contact dell to sort out this mess and go through the repair/exchange process. however, if you are up for it, you can try the steps detailed in post #1.

I had another skype call yesterday with one of the viewers referred to me by stephen from ownordisown youtube channel. He had 2 laptop replacements both having the same stuttering issue. He ended up fixing it himself with thermal pads adjustments (though he did went a bit overboard on the thermal pad thickness) from ownordisown's youtube video and the stuttering was gone. one thing to note, he also said this is something challenging, hard to get right and is highly dependent on your skill level. he is still tweaking the thickness from our conversation. Bright young chap! happy for him.

 

from your GPUz graph, you are experiencing the same stuttering symptoms as @prestona . GTX1080 is known to power throttle and thermal throttle when it reached a certain limit. So the best explanation is that when you lifted the back of the laptop, more air is drawn in by the fan which resulted in much cooler GPU. however, once the GPU is cooled enough, any power/thermal throttling limit is lifted allowing the GPU to draw even more power, thus stressing the power delivery components more causing them to overheat. and because the VRM, chokes, GDDR5x are all underneath the heatsink and thermal pads, they dont receive much cooling benefits from raising the laptop's back.

You have 2 options now. 1, the recommended one, is to contact dell to sort out this mess and try your luck with another unit.

2. if you are skilled and have experience in repasting cpu & gpu, it's worth for you to try checking the thermal pads. pay special attention to the tiny square black VRMs and GDDR5x. check for thin film plastic alienware assembly line forgot to remove (this was shared with me from one of my skype calls) on the thermal pads, and check for thermal pad contacts with GDDR5x and power delivery components. Refer to post #1. be warned this requires high level attention to details, patience and skill. not everyone can get it right the first time.

 

Thanks alexnvidia for the info

I have reviewed this thread pretty thoroughly (also watched the YT vid I think you cameo'd in on this topic), and a couple other posts about repasting/replacing thermal pads. I do have experience with adding thermal paste on a laptop (used to own a Gateway FX with a 9800m GTS which I added Arctic Silver to), and I have built a couple gaming PCs from scratch, so I know that with some time I could get to the bottom of this issue.

Unfortunately I bought my r4 right before I leaving the united states for a few months, and where I am living now there is not a reliable way to mail things into and out of the states, so returning the laptop is not an easy option for me. Also, getting the necessary materials to do a repaste job isn't easy where I'm located. So I'm looking for any physically non-invasive options I have.

I happen to have a laptop cooling stand (something like this http://a.co/7HW4C8z), and I can share that when I've tried to reproduce the issue I haven't been able to do so while the laptop was on this stand. Since you have a good idea of the layout of the heatsinks and chassis, do you think its possible that when the laptop has an entire surface to rest its chassis on, rather than being propped up and supported by only a point at its very rear, that the chassis applies the necessary pressure to improve the contact between the heat dispersion components? Could that be why I have had trouble reproducing the issue on the laptop stand? What are your thoughts?

Also, I am willing to sacrifice performance to resolve this issue for now. I'm thinking of using the older BIOS, but do you think that undervolting the GPU could help in this situation? (I have some experience undervolting as well).

 

I totally understand how difficult it can be for those not living in the states and not having easy access to good thermal pads and easy return policy. Took 1 month to receive fujipoly thermal pads i ordered from USA to reach my country with a shipping cost alone near $50. and because like you, im not living in the country i bought this laptop, exchange is a great big hassle, took over 30 emails back and forth with the indian tech support and over a month before i can get my hands on the replacement. so i figured, screw it, i'm fixing this mess myself.

Having said that, placing a laptop cooling stand with good airflow does help. in fact, if you removed the entire back cover of the laptop and rest it on a laptop cooling stand with active fan cooling, the stutter will go away as demonstrated by @Zer0K if i remembered correctly. this goes to show the power delivery components, GDDR5x are actually not cooled properly and overheating. and if you take note your PCH temperature, by sitting the laptop on a laptop cooler, the PCH temps dropped a good 10C or more.

As for the argument of laptop laying flat or sitting on the cooler, it doesn't make much of a difference in terms of pressure & weight distribution. I have experimented several times adding thick rubber dampers on key locations between the heatsink and with the mid frame screwed down, hoping to provide extra pressure to press down the heatsink with the laptop's own body weight. so much so to some point the entire mid frame got deformed but still i get no noticeable improvements.

I myself am using cooler master U3 plus. and i have tested my fixed laptop with both the cooling fans on and off for over 2 hours gaming BF1. with the fans off my PCH will hit 80+C and with the fans on, PCH will hover around 65C. in both situations, 100% stutter free.

 

indian tech support is like communism

the only difference is that communism actually works and accomplishes something

 

Best non invasive option right now is the old vBIOS. that should get things working and temps down. Won't be a noticeable performance hit either.

There's been a lot of talk and a lot of us got lemons but there's no clear way to solve it. Proper paste and pads can fix it but it won't work miracles if your components have already failed in some way just by using them until now.
It's a fairly big job, takes some time and multiple tries so you should only do it as a last resort, if it still stutters with the old vBIOS.
The cooling pad is a good idea, obviously. If and when you can, run it.

If you are outside the states, but in a country where Dell also operates I would start some sort of support process, just to get things going, there will probably be a while before they get any parts but you get to keep your PC and the issue will already be registered with Dell. Inside the return period. So if they can't or take a long time to fix it, maybe you can still return it for a new unit when you get back, outside the usual return period.

 

one more thing to add, i don't think there's an easy way to undervolt laptop GPU. unlike desktop GTX1080, power limit, voltage etc are all locked for laptop. you can however undervolt the CPU. quite a number of people including myself have had success with the old vBIOS which limits the power delivered to the GPU to 150W. you can try that if you do not wish to get your hands dirty. but i wouldnt continue using it for long period of time especially when the underlying problem is still there. i hope you bought extended warranty as well.

 

Happy to report I received my refurb in mint condition - 7820HK + 1080 GTX.
Added some RAM and an SSD.

No stuttering whatsoever; played some BF1 intensively (ultra + taa + 150% original resolution at 1440p).
Full hour went buy with no stuttering.

Played some Attila and TW Warhammer afterwards.
Might be fixed by Dell already?

I have already prepurchased Kryonaut paste + expensive fujipoly 14k/w pads.
Oh well, that'll come in handy once I repaste in a year.

 

Its possible to undervolt aw17r4 gpu with afterburner und custom curve (ctr f). I do it and it is working great. Voltage does not go above 1.0v at 2GHz. No stutter.

Gesendet von meinem SM-G955F mit Tapatalk

 

sounds interesting. i wasn't aware of this until you mentioned it. care to share a bit more on this? an example or a screen capture of your voltage curve would be great. and how do you achieve undervolt with this method?

 

waaat?

 

How did you get a replacement?

Did they try on site repairs first?

 

This mess will never be fixed if not Dell use quality components, quality pads + properly QC on the job or push out firmware who decrease the power due use of weak power delivery components. This is like the lottery. You will most likely loose as always!! The overheating mess has been there now for almost 9 months. Soon is this laptop revision EOL and Dell will push out new models without fixing current models!!

 

Everything you need to know is explained in this video:


However, I never experienced stutter, not even at stock. I repasted and repadded according to @iunlock and your guide, put heatsink on PCH and SSD and use the OC vbios. By using custom curve in afterburner and undervolting the CPU and elevating the notebook I get very good temps for CPU and GPU, no throttling, no stutter.

When set up correctly and put a little work into this is a very nice notebook. I guess I will soon sell my Clevo DM3 in favour of this one, SLI for gaming seems to be a dead end.

 

so it only works with beta version of afterburner?

 

No, in the meantime it is not beta anymore. Just download newest version of afterburner;-)