*OFFICIAL* Alienware 17 R5 Owner's Lounge

yes

 

I never had any luck flashing a modded bios with the Aptio / Intel tools on my 17R4. Probably easier to buy an SPI programmer, a few spare bios chips, and try to solder a socket for the bios chip so you can easily flash and swap. Obviously would void the warranty so probably not a good idea on a $3000 laptop..

It seemed like it may be possible to update hidden bios settings via the grub efi shell utility "setvar" but that utility has a weird bug on the 17R4 where two setup regions are detected and it limits the amount of accessible settings. All that I really wanted was a way to change the eDP MUX from CPU / GPU while using Linux. If you want to try this you will need to use UEFITool to dump the setup IFR region of your bios, so you can see the offsets for each bios option and possible values. This method has been used successfully on a few other Dell models like the XPS 15, so I feel like it should be possible with further experimentation.

 

If anyone could please diagnose this: HWinfo: IA PROCHOT and XTU: Thermal throttling?

The CPU is not that hot at 64C so what can it be?

 

Got to look at all temps bud. Your pch is scorching hot at 99c. Prochot gets triggered when a component gets too hot. I believe 85c triggers prochot.

 

Yeah, He is about to cook his PCH chips

 

Clearly I'm not understanding something here, sorry if this has been covered already. My i9, if I run Intel extreme utility CPU stress after about 30 seconds my clock drops from 4.3ghz to about 3.6gh and I get Power Limit Throttling . . . . what am i doing wrong?

 

Lol you think that's cooking it, I got 102C on my current (maybe soon to be old depending on AW's escalation) machine's PCH

 

Your’s has passed the boiling point and instead started it’s way to be fried

You need to raise the Amps and Power limits.

 

Your thermal throttle seems to be due to PCH because it controls the CPU I think. I did not know your PCH reaches 99C after your PCH mod. Looks like your mod did not work. I would recommend to do it again. Maybe the copper heatsink is not tightly stucked on PCH. Did you use a thermal pad between the shim and the PCH or not? I used a 0.5mm (maybe 1.5mm I cant remember) gelid pad and a 1.5mm thick Copper Shim 20mm x 20mm Heatsink over the thermal pad. As described, I used 3M tape to cover the PCH surface and then an instant glue to stick the 4 sides of the copper shim on the 3M tape on the PCH. I put pressure with my finger until the glue dries on all 4 sides to make sure it will have proper tight contact with the thermal pad and the thermal pad with the PCH. I hope that helps.

 

So, I just repasted my system with TGC. But, sadly the difference between Kyronaut wasn't really big. Got a maximum of CPU: 95 C and GPU: 75C . Though, this system was being tested in a really hot room (prob 34-35C, sunny hot day here in Singapore). But, the system didn't thermal throttle or skipped a beat. So I will say, pretty impressive and very reasonable.

Hopefully my core differential remains 3-5 C as what I am experiencing now. Also, I nearly spilled LM all over my board haha

Thanks @doofus99, @alexnvidia for the tips!

 

For certain this is my third PCH attempt and it does not work at all. I need to invent another method of cooling the PCH.

However the thermal throttle is not due to the PCH because it also happens when it is at 93C-94C.

 

Hello everyone. I'd really hate to ask but I'm getting my 17 r5 today and I opt to get the basic drives to save money but intended on getting a samsung 970 m2 ssd 1tb and a firecuda for storage. How would I go about properly installing the system all over again as once I swap drives the recovery software will be on the older drives. Would I just be able to install the software via flash drive? And what do I specifically need to enable in the mb for the ssd? I'm actually quite out of practice so I embarrassingly need to ask :x.

 

Power limits are locked in bios... And I don't know where to change amps... But last week it was not power throttling and as far as I know nothing changed. Last week it was holding 4.3 and power limits were still locked with the new bios... It's gotta be something else?

 

Nevermind I got it figured out, im back up to 4.3ghz without having to touch any manual amp or power limit settings.

 

If I am not wrong, Dell told you that the thermal throttle starts at 93C?

 

I personally think it's your gpu vrm. With my 1060 in my 13 r3 with its ~80w limit can get prochot. Considering your gpu can push way more watts would increase the temp even more of the vrm. So try to undervolt the gpu and see if you get prochot still. This will also bring your pch temps down. Download msi afterburner and undervolt then tell us your results.

 

So what was the problem?

 

I have done it on my Samsung EVO 960 OS drive (check my signature). A number of options:

1) Best solution: Do clean install downloading windows 10 home/pro (there are many tools to download or use microsoft media creation tool to create an installation usb drive from current system). OS is going to activate automatically when you login with your microsoft account. Then install all the drivers from Dell (specifically for your service tag).
2) Use a tool like EaseUS Partition Master 12.9. Use clone disk option or transfer OS to SSD (new Samsung SSD has to be at the position 2). It makes a copy of whole drive or system but it may or not have precisely copied everything.
3) create recovery usb disk from current system and use it to install on new drive? I have not done this though

I have my original disk too but not inside the laptop. DO NOT FORMAT the original disk because you are going to lose your factory restore (factory reset does not work from new drive even if you copy the restore partition on new drive).

For samsung NVME drives, you also gonna need a heatsink for sure (I had to make a mod for one of the heatsinks as well because of limited space) but temps can go high (in the high 70s when using a good heatsink) even with that when gaming for long. Also keep the system on RAID. It reduces the samsung NVME drive speeds (still fast about 2700MB/sec read) and temps or you gonna have the system lagging when playing heavy games. On AHCI mode, it is using much more speed (3000mb/sec read) but temps can get out of control (in the 90s range) and the system may crash only because the drive stops responding.

Motherboard recognised both my 960 drives (even though they are PM and not SM) when I installed windows from my windows installation usb drive.

 

Tried disable "BD PROCHOT" through ThrottleStop? If not Dell really have screwed up completely this time, disabling (uncheck) BD PROCHOT in ThrottleStop main windows should prevents other items on your motherboard from throttling your CPU. However, do not run TS simultaneously with XTU services running.

Edit.You could force max fans with Hwinfo to see if it changes the behaviour. Follow temp and see when the throttling kick in.

Even your ssd have passed the throttling temp with light load ( PCH temp goes hand in hand with the SSD's).

Not much have really changed from previous gen models with Skylake and Kabylake.

 

You cannot control any fans with Hwinfo. Regarding SSDs, I would use a heatsink but this is high for default (slow) SSD.

 

You should be able to. All previous gen Aw models could do that due compatibility with Dells EC. Google it. Or check it out/post in Hwinfo forum.

 

Well I have tried. The fans menu does not even show up when I press the fan icon on Hwinfo (EC support is enabled)

 

Tried disable Dells ACC software and see if it interferes with Hwinfo64 fan control? Hwinfo64 have been the only saviour for the Alienwares the last 6 years (all had crippled fan profiles - Fans started spin up all too late and with no possibility for max fan speed - Fans reached max speed only when you flashed bios). Oh'well, Dell has really screwed up this model revision. Because Hwinfo fan control historically have always worked with Dell's EC firmware. Tried Speed fan software?

 

hey i got it. My new alienware
once again:
i9
gtx1080
uhd
32 gb ram 2666mhz
win10pro

at first: bios version is 1.1.6!!! That should be good
I think all the alienware are packed up very bad! I had seen a lot of shipping companys and how they throw packages cross their car and so on...every bigger vibration can damage the heatsink or create a gap between heatsink and cpu, gpu etc. especially when i read what you all wrote about the quality inside the alienware...
But for my first impression its an amazing notebook
At first i will create an full image that i always have this original state. Maybe with acronis? or windows 10? What do you think? And why dell didn´t send a recovery disk or somthing else? Can i download that? Or how could i switch back to original state if i want that? create image as i wrote up there?
Other question is why i have 4 or 5 partitons like recovery but all of them are empty?no reason why...
and what about farcry5? normaly it is includet with this notebook...also downloading? but where...
questions questions questions....sorry

in the next days i will check all the temps and the other things and keep you informed!

 

I do not want to uninstall ACC now that I got it to work properly. Maybe someone else can try.
It is true that I never heard the fans go so fast as during the bios. Fans respond late to temps (going faster only when average temps are going higher during gaming) and this is why we see temp spikes anyway. Because they last for 1 sec that the fans do not even have time to respond.

 

After a whole day's experimentation the situation is bad but I have learned a few things.

1) The PCH reports 94C in Hwinfo, but I have consistently (for many days) measured it to be at about 72C. Today I used two electronic thermometers and one IR, all showing 72-73C when Hwinfo shows 94C+. Either Hwinfo is getting it wrong or the BIOS is getting it wrong (because the HWinfo author implied that he uses some BIOS service to access the PCH - still awaiting clarification). Physically however we are 20C less than what's being reported.

2) Due to only 3 screws on the CPU side, the left hand side of the CPU receives less tension than the right hand side. I have discovered that the screws have to be screwed top first and bottom later and ignore the numbering, else you get the left hand side of the CPU not depressed enough and very hot!

3) I am still getting the thermal throttles, but only when the GPU is at 180W and when *everything* is at around 70C-80C. So with that in mind I took the thermometers out again and measured and believe have found the culprit. It is two ICs situated below the GPU and with the GPU heatsink's extensions over them. It hits over 104C, throttle happens, drops to around 95C and the cyclerepeats. I do not know what these ICs do. I re-padded again, made sure I had solid contact, but those ICs are cooking. This all started when I decided to repad! I should have left it alone because it never did this with the original pads.

 

Yeah, make it ready for RMA

 

Oh funny the vrm's are too hot XD. It looks like you stacked the thermal pads. Am I wrong? You lose a lot of the effectiveness of the thermal pads when they're stacked.

 

This weird thermal throttle did not exist until I changed all the thermal pads. Obviously my fault. That and the rook temps now over 32C!

The ICs in the photo with 100C are at the extremities of the heatsink - I am not sure if they are voltage regulators. I did not take a close up photo. I have re-tried with various combinations of pads to make sure there is good contact and there is.

Maybe my Arctic pads are not that good and need to buy better.

​

 

So to my surprise upon receiving my 17 R5 it came with a Samsung nvme m2 ssd preinstalled. From what I've read on the forums most people did not get a samsung SSD. It isnt an evo or pro but from the specs it does quite well. I may just keep it as my boot drive and install the extra 970 evo m2 for gaming and use my firecuda for storage. I'll post some benchmarks when I get a chance!

In case anyone is wondering it's a
Samsung NVMe M.2 PM981 256GB

 

Dont worry this has happened to me (remember the high temps immediately when I replaced the thermal pads. I can see on the pics you use two thermal pads to cover the distance between the VRMs and the heatsink? This is not right. Use a 1.5mm ones. I use gelid pads. I also use K5 pro which replaced the white thermal pads. You can actually use it to replace all thermal pads but I use it only instead of the white thermal pads.

Which samsung model is it?

 

Samsung NVMe M.2 PM981 256GB

 

Can you post a pic of its heatsink? PM981 is fast. Also I wonder how its temps are going to be during gaming. In addition is your bios using RAID mode?

 

He never said it had a heatsink? I'm not sure if alienware ever used heatsinks on the ssd. Someone correct me if I'm wrong and post a picture of it if there is one

 

Hhmm I had a lot of trouble with this alien command center and the light fx. The alienhead in front of the keys have had no light anymore. After complete reset and disconnect battery it light up again but to change the color it is nearly impossible...but it looks like it is the acc!
@Aristotelhs2060 :you wrote a lot of problems because your cpu trop down to 800mhz in idle. Hoch can adjust it in energy options. There is standard 10% at cpu performance. Increase it to 100 and your cpus will never trop under 2800mhz.

Temperature in idle at 4.5ghz: about 50
At 800mhz 45
No cpu usage at all no oc.
Next test is stress test.
Fan adjustment is really locked in hwinfo. Also after shut down all acc processes. Funny. Also no rpm of fans in hwinfo...

 

ACC is really, really buggy. I uninstalled it, then re-installed it, then it said that it was downloading updates and modules, which it did, I setup all the lights, and have not touched it again. If it weren't for the lights, I would not even install it.

 

ACC is full of bugs. On mine, sometimes it shows only the Fusion and the library menus without any reason and I have to reinstall using modify option to make all the menus show up again.

Which option are you talking about for the CPU? I think I have tried everything. But, is something is running, it never drops to 800Mhz. Only during idle.

 

I cannot think how such a fast drive can operate within acceptable temp ranges without a heatsink. And yes I heard that Dell is using heatsink for such high speed NVME drives (talking about over 2000mb/sec read speed and not the ones with 500mb/sec and the operating temps rise accordingly). I had to use heatsink on my PM961s anyway

 

I agree Samsung ssd's need heatsinks. Well if you come across alienwares oem ssd heatsinks let me know. I've never seen one.

 

You have to Search for energy options (don’t know the right translation in English).
In this new window go to more energy options like the picture and then search in new window for cpu energy. There you can increase the minimum power of your cpu. It works I already checked that.



Hmmm why is this acc so buggy...bad.
That means better no oc,uv with acc?

 

Did you ever had problems with your touch pad? 3 times it freeze and I could not move my mouse...I had to restart computer with error message...

 

No problems with my Touchpad here and no 800mhz dips.

 

Maybe your energy settings for min cpu performance are at 100%. I also set it now to 100% minimum cpu performance and 100% at maximum cpu performance. That means my cpu frequency is between 2800mhz and 4800mhz depend on cpu usage.
And then it never drops below 2800mhz.

Maybe I have problems with some software and touch pad...I don’t know sometimes my mouse also jump from one to another point...

 

Tried to reinstall the drivers? And yes i have 100% in the energy settings and my CPU never Drops below the target clock speed (24/7 4.3ghz)

 

Hello,
Does anybody have a link to the installation order for the drivers.
I am in need of a clean install of my new Alienware 17 R5,

Best Regards.

 

which bios and which windows version are you using?

Start from all intel drivers (first chipset drivers then the rest) first, then wifi ethernet bluetooth and finally dell specific drivers. This is what I did for clean install and I think this the order you should always have. Always restart when asked to.

 

bios: 1.1.6
win 10 pro

but not the newest

 

Sorry for replying late. Taking summer college courses so ive been busy. The SSD did not come with a heatsink whatsoever. I will say this though I'm not too happy with performances so far or maybe I'm just simply jumping the gun and being spoiled by others doing the same benchmarking and scoring higher than my AW17 R5. My current model is the 17 R5 2k panel i9-8950HK 1080 GTX. Currently using stock settings with the 1.21 bios. My cinebench is really inconsistent scoring anywhere from 981-1100. I dont entirely know if these scores are good per say with this processor. I will say this... When I've watched others post stock videos they score 1200-1250 and dont overclock or claimed they haven't overclocked and its right out of the box. My stock settings using HWmonitor pro shows my 4666 and temps range from 78-87C. When I use the OC settings from either 4.8-5GHz I actually score lower than I do with stock settings. On stock with fan profile set to performance my highest score is 1196. If I am simply being spoiled by others scores, let me know. I haven't tested any games yet as I've been busy but only used Cinebench and Time Spy. My Time Spy score using default settings was 17206.

 

There's something wrong there.
Maybe try and monitor hwinfo when you run Cinebench.
Ive got the 8750 and scores of 1000-1150 are normal... Once hit 1200 ( must be a fluke) with undervolting and cooling.
The i9 should do and can do better.

 

Bios 1.6.1 and not latest windows. One of those two is causing the drop to 800Mhz indepedently of energy profile settings. Hmmm it makes sense now