All CPU and GPU OC differently. It's more like a draw of luck, silicon lottery if it makes sense. That said,
For CPU, if you are using 7820HK, you can OC all the way to 4.4GHz on default voltage quite easily and still remain stable. Pushing it beyond that speed requires additional voltage. But for praticality, 4.1GHz is where the sweet spot is. You can set both PL1/2 to 100W, set all four cores to 4.1GHz and using XTU or throttlestop, undervolt -100mV.
For GPU, you can use MSI Afterburner and start with +100MHz all the way to +200MHz if your GPU is willing. i find the sweet spot around +150MHz. GPU GDDR5X can easily net you 11Gbps or +500MHz. Then again, your milage may vary.
To keep OC temps under control, please ensure you have repasted and preferably repad before you OC.
Enjoy and good luck.
-
-
Unfortunately and as sad as it sounds, i have to agree with you on this. Even laptop build as recent at end of June with Sunon heatsink still have a massive over 20C CPU core temperature differential with the hottest core reaching over 95C and even after repaste, the cpu core temperature differential remains at about 15C between hottest and coolest cores. Seriously pathetic.
-
Thanks alot, but i have totally no idea how to diy for that, is there anyone can do for it in Kuala Lumpur ? And which brand of thermal pad and thermal paste should buy ? I also having the issue same like you. i have played overwatch for around 10-15 miniute and it start to blink blink and i check my gpu-z it can go up to 91c. @alexnvidia
-
and a last question of mine is will the alien ware void our warranty when we do the repaste ?
-
Thermal Grizzly Kryonaut is good. Should be easy to find in KL. GPU temp shooting up to 91C is TOTALLY not normal. you need to check your heatsink and gpu contact. even with the stock thermal STAMP, you will still be looking at under 70C for GPU temp under intense gaming. if you dont know how to DIY, contact dell and let them resolve it for uVasudev likes this.
-
no. That's the good thing about alienware. unlike MSI and other brands, no warranty BS sticker exist. Heck alienware even encourage you to open up your system yourself with all the video tutorial you can find online, service manual etc.Vasudev likes this.
-
Thanks alot mate. Can you show me your XTU setting in picture cause im quite a newbie or throttlestop also can. as long as which one you using. @alexnvidia
-
what CPU are you using?
-
mine totally same spec with yours. alienware 17 r4, intel core i7 7820hk, gtx 1080 and qhd one @alexnvidia
-
-
Having GPU temps at 91C is totally not advisable to OC. But if you insist, you can go to BIOS, enable CPU performance mode, enable customisation, set core 1-4 as 4.1GHz, Power Limit 1 & 2 to 100000.
Then open XTU, go to advanced tuning on the left side panel, look for CORE VOLATAGE OFFSET, reduce it to -0.100V. That's all you need to do for CPU OC.Vasudev likes this. -
OK, I managed to get one of these supposedly fantastic beasts. I am trying to deal with Dell support to rectify the issue but it is not yielding any results.
They are telling me that the thermal results are correct and I have posted over and over again different images of my results. No thermal throttling occurring but when I am playing anything, online or offline, I get the kinds of freezes described here. Below is the pictures I have for various games and the afterburner showing it is not throttling.
http://imgur.com/a/aiHJm -
Is it true that some have resolved the stutter by laying the laptop flat on a desk rather than having the back raised and/or using a cooling pad? Something about an airflow issue?
I haven't tried it myself, but curious. On 1.0.8 atm w/ no stutter, but not happy with the downgrade.... -
Highly unlikely. Laying laptop on a flat surface without any gap will only decrease the airflow and raise overall temps.
-
Hotter graphics will decrease the performance. Hence Less pressure on the VRM's etc. Pascal will loose 13MHz clock speed for every 5C degrees temp will increase over 46C.
-
@Vasudev did the bios update and everything seems to be working now. no stutters anymore when playing doom and dota2. I asked for a mobo replacement a few hours before testing and they will be sending me a new mobo now my question is should i still go through with the new mobo or keep my used mobo? I also asked for a battery replacement since mine shows 12% wear level via hwmonitor. To be fair the dells support in my country is superb! upon talking to a pro support agent i told him i was not happy with what i am experiencing and asked for my permission to remote access my aw and i said ok and put him on speaker while watching the latest ep of GoT. Parts+tech should be coming before the end of the week.
Also they will be sending me new fans just incase its a heating issue he said so yey new fan
Vasudev likes this. -
If everything's working fine, then you could replace only the battery. I think you're getting the KabyLake CPU based board, so your single core boost will be 3.8GHz instead of 3.5GHz with better thermals and hardware support.
Be careful about disassembly check everything in ePSA quick test to see if everything is fine even battery wear level, AC adapter wattage etc.jpsm likes this. -
i highly doubt that i will be getting a kabylake replacement since the replacement parts come from china i am pretty sure they will have tons of replacement skylake boards but of course if they are giving me a kabylake that would be great! Hope i wont need to do a reinstall because its such a hassle to configure it to my liking lol. Hope the battery will be better this time and i think ill go through with the new fans
Vasudev likes this. -
So i installed the new bios today with Alienware tech support. Also installed latest nvidia drivers.
The stutter is not occurring on the latest bios. 1.1.8
They probably reduced some of the wattage going to the gpu. But it's something for folks to try. Atleast you wont have to use the old 1.0.8 Bios anymore if this helps others as well.jpsm likes this. -
A day before the tech came in i played tons of games and realized that the stutter was still there and now happened more often! Despite me posting that everything was ok after the vbios/bios update. I am reporting today after a day of using my laptop extensively. I had my fans, battery and mobo replaced. Now battery shows up at 2% wear from the 12%. I decided to do a repaste also with gelid extreme. I am very pleased with the results now! I hope the stutter does not come back any time soon though! My temps after gaming for 5-6 hours. -140mv cpu undervolt.
Last edited: Jul 27, 2017Vasudev likes this. -
it will come back like mine did.... If it does not then lucky for you! if you still have warranty i suggest asking for a new mobo until you get lucky
-
God I need to undervolt mine....
Is there anyway to force the undervolt all the time? Everytime I turn the PC on.
Or do I still need to open XTU manually everytaim? -
pathfindercod Notebook Virtuoso
*This is a copy and paste from another member*
I found this in my personal notes I keep when I find stuff on the net. I can't find the original posters thread here. All credits go to him!
Background: XTU is unpredictable when undervolting. It is supposed to stay after undervolt profile is made, but this doesn't really work all the time. About 50% of the time I find I have to re undervolt my profile because it was lost or "quit unexpectedly". I got annoyed that I would have to launch XTU (or HWinfo64) every boot/log in to make sure it is working as intended
After much trial and error I have been able to undervolt in XTU without having to launch XTU every boot
Note: this allows Powershell to change/set core voltage without needing to start XTU at all
WARNING: only do this with a core offset that is RELATIVELY stable (or add a delay to task scheduler)
really easy to determine relative stability (i.e. it doesn't crash your computer right away, so you can change it or stop the scheduled task if you need to)
Just: Undervolt in XTU and run OOCT for 5 minutes
Or better yet, just use an undervolt that you have tested and used extensively
DIRECTIONS:
1) Copy this code to a notepad:
$status = get-service -name "XTU3SERVICE" | Select-Object {$_.status} | format-wide
if ($status -ne "Running") { start-service -name "XTU3SERVICE"}
& 'C:\Program Files (x86)\Intel\Intel(R) Extreme Tuning Utility\Client\XTUCli.exe' -t -id 34 -v - YOUR VOLTAGE OFFSET HERE
sleep 4
stop-process -id $PID -force
*Your XTU file path may be different*
EXAMPLE -0.150 offset (for clarity because you don't use decimal place in code)
$status = get-service -name "XTU3SERVICE" | Select-Object {$_.status} | format-wide
if ($status -ne "Running") { start-service -name "XTU3SERVICE"}
& 'C:\Program Files (x86)\Intel\Intel(R) Extreme Tuning Utility\Client\XTUCli.exe' -t -id 34 -v -150
sleep 4
stop-process -id $PID -force
2) Save file as "filename.ps1" to a known safe location (I saved as undervolt.ps1 and saved in a folder called boot in C:\)
3) run the file to be positive it works in powershell. I.e. set an undervolt of 0 in xtu, then run the file in powershell to make sure it changes
4) Search for powershell, shift-right click, run as administrator and run: Set-ExecutionPolicy RemoteSigned
5) Copy this code into notepad:
command = "powershell.exe -nologo -command C:\ Your .ps1 File Path Here
set shell = CreateObject("WScript.Shell")
shell.Run command,0
6) Save as undervolt.vbs
7) You can run using Task Scheduler. Simply open task scheduler and create a basic new task. Your triggers are going to be, At log on (or whenever you'd like) but at log in seems to work best
Actions: For this brouse to your undervolt.vbs it is a script so no program needs to be run
Check "Run with highest privileges" in finish step
Note: the .vbs script allows powershell to run in the background without a powershell window popping up for a few seconds.
If you'd rather the powershell window pop up, you set action program as powershell.exe and put the .ps1 file path under arguments
The power shell window will then pop up for 2 seconds and show the undervolt -
Whenever i turn on my pc its undervolted already. Didnt do any mod to get it to run on startup. I think i got lucky with the replacement heatsink i believe it has better contact with the cpu and gpu thats why i got nice temps!
Sent from my SM-G935F using Tapatalk -
pathfindercod Notebook Virtuoso
How's that? What software are you using? -
Xtu
Sent from my SM-G935F using Tapatalk -
Try ThrottleStop after trying out optimal undervolts in XTU, everyone does that. Moreover, XTU increases CPU temps by 5C while TS doesn't.
bennyg likes this. -
pathfindercod Notebook Virtuoso
Xtu does not automatically start and underdervolt on bootup. You have to manually start it unless you pretty much do the above steps.Last edited: Jul 27, 2017 -
When i open hwmonitor after bootup it shows -140 undervolt. Idk. Ill try to observe later. Ill also do the steps just incase hwmonitor is wrong
Sent from my SM-G935F using Tapatalk -
It's mentioned to NOT reuse fujipoly pads... but the directions on the first page say to put the heatsink back on, then take it off to check for imprints on the fujipoly to make sure they are making contact. Do I have to go through 2 sets of fujipoly pads here? Or is "reusing" in this case fine? These pads are certainly not cheap and I would hate to use up a whole strip just for testing if the heatsink is touching it.
-
It's ok to use the same pads, just don't crunch them and you'll be absolutely fine.nemoris likes this.
-
F*ck me sideways.
I'm having freezing again...
It's been a really hot summer here in Portugal, my unit has been mostly working flawlessly since I got a new motherboard and heatsink, but I'm starting to notice some freezing coming back.
Mostly on Battlegrounds, after a few games, everything gets quite hot. CPU and GPU hover at around 80 degrees but the whole laptop itself is pretty warm.
I have conductonaut and arctic pads, everything was making good contact. I'm on latest BIOS and drivers. Same thing happened with the penultimate BIOS version.
Two weird things I have noticed with my unit:
-When I first repasted with Conductonaut I also did the PCH mod. I got a really good Raspberry Pi heatsink which had lots of surface area and was made out of aluminium. With this heatsink installed I still had freezing. I don't quite remember the PCH temperatures though. Removing the heatsink made everything work fine again.
-I'm not too sure about this but I think that when I have the laptop level on my table, nothing propping it up what-so-ever, it works fine, no freezing! But if I do prop it up the freezing comes back.
Never in my life I have seen such a weird behaving piece of tech... It doesn't make sense. I think the PCH had no heatsink for a reason, maybe it's doing some kind of automatic thermal throttling or voltage regulation. Keeping it cool makes the CPU fail to 800MHz at times...jpsm likes this. -
Just a heads up to the folks in the thread, I'm not sure if this has been discussed before but are you guys SURE it isn't being caused by GFE? There is a well documented bug with Optimus because NVidia drivers are garbage and I've seen the same thing myself. The only way to really "fix" it is to remove GFE and remove all drivers completely from the computer and reinstall using ONLY the Dell/Windows Update drivers. Using the current drivers+betas bring the freezing back unless you disable the iGPU.
-
What about in GTX1080 units that don't use the iGPU? Nor Optimus?
I've seen GFE working it's magic. Just yesterday I tried to record some gameplay and everything got messed up.
Is there no alternative to installing Dell drivers? what about the latest Nvidia ones and just not selecting GFE?
I'll try that later. -
Well *all* of us use the iGPU is what I'm trying to say. The way hybrid graphics is routed is through the PCH and through the iGPU to get to the display. The only way to get a straight feed off the dGPU is from the DP or HDMI port on the rear. As for alternative drivers beats me! I've been banging my head on this one for days now and that's the only one that works. I wish GFE wasn't so ****ty because I would honestly like using Whispermode and the other features but it just made my 13r3 run like ****.Shark00n likes this.
-
Does anyone know if GFE is known to cause issues with MSI's Afterburner or other monitoring software? I've had games simply not wanting to start if I start monitoring
-
I've used that for years and never seen an issue like you describe on both NVidia and AMD
-
RTSS interferes with some apps and is known problem that wrecks havoc while on battery. Maybe its EC issue. When you're using GFE's monitoring service avoid using RTSS or close GFE and use RTSS.Shark00n likes this.
-
Actually it does. It creates a windows service with a Delayed Auto Start. This is why some people think the XTU setting has not sticked. You gotta wait a few minutes for the undervolt to get set after boot up.Vasudev likes this.
-
Try to get a replacement mobo bro. I was lucky enough to get a good heatsink w good contact w my cpu ang gpu. My temps at -140undervolt wont go over 70c even under full load. I no longet experience freezes after getting a new mobo too. I have busted speakers now though and will be getting a tech coming to my place hopefully today or monday. If i get 1 more problem with this machine i am selling this already. I owned a 15r1 and had exactly the same problems! 94c on cpu and busted speakers after 3 months. Thank god dell has home service/on site warranty so no hassle on my part except the 1-2 hours i waste watching the tech service my machine. Hope you can get a replacement bro tbh this should not be happening to any us since it comes with a hefty price tag...
Sent from my SM-G935F using Tapatalk -
I have the same problem with my aw17r4 and have had a heatsink replaced but my cpu temps still go upto 98 degrees however since the heatsink replacement and vbios downgrade the stutters aren't as frequent i was able to play BF1 for 40 mins and Witcher 3 for an hour straight. But a few days ago it started to stutter again within 16 mins of playing witcher 3 i also had hwinfo running in the background and apparently 2 of the 4 cores of the cpu throttled during gameplay with a max temp of 98 degrees , i called the support again and they are having the heatsink replaced yet again.
The question i wanted to ask is should i stack the pads from the old heatsink on top of the pad on the new one, since I can't find any good quality pads here in India? -
It depends on what thermal pads you have currently on your heatsink. those fujipoly 11 and above are not meant to be transfered once used. They crack easily, they dont return to shape. pretty much useless if u ask me. and once compressed, the thickness has totally changed and might not give good contacts on another heatsink. stacking on the other hand might cause the pads to be too thick and raised the heatsink causing poor GPU and CPU contacts. 98C is way too hot,
worst case scenario, just use back alienware stock thermal pads that come with the new heatsink assembly.c69k and Altairwarz like this. -
It's my third mobo. I really don't want to go thru that again... Already had to take a day off for them to fix it the second time. I could have spent that day at the beach............
I'm on my second heatsink too. A Sunon one. Temps are not so bad but the whole chassis sure gets toasty. Portuguese summer is a b*tch to computers, pretty much all gaming laptops have issues around this time.
My old GS70 would game at 92ºC and the whole chassis was freakin hot! I swear you could fry an egg on it...
I uninstalled GFE and nvidia drivers and got the ones from Dell. Things seem to be back to normal with no freezing
Vasudev likes this. -
Scratch that. Freezing full swing.
It's not 100% like my first ones. Games dont fully stop this time. But FPS counts get down to 10 and it's super jittery.
stock cpu was getting to 90 and gpu to 85
with -120mV cpu is at 82, gpu 85
both throttle way down to 800 and 130mhz respectively
I have conductonaut and better pads, ambient temps where I live are a bit hot but it should work fine...
I'm gonna try the latest BIOS. -
Try removing undervolt
-
Are thermal grizzly thermal pad any good?
-
i find the arctic thermal pad recommended by @Mobius 1 to be really good compromise and reasonable performance
-
Do you have MSI AB and RTSS running?
-
Sometimes I did yeah.
But it still happens when I only run HW Monitor
Alienware 17 R4 GTX1080 in game random Stutter & Freeze
Discussion in '2015+ Alienware 13 / 15 / 17' started by alexnvidia, May 6, 2017.