Hello, I recently purchased an AW15 R4 with a 8750H and a 1070 GPU, I was quick to update all prompted Bios and software updates on the support assist app as soon as I got it set up. During gameplay the cpu stays well above 95° and I'm scared about longevity and other issues. Its really been a huge disappointment to have to underclock and undervolt both the CPU and the GPU to get below the 85° mark ( still hits the occasional 90° ) Please help me out, I have a strange feeling it has something to do with the GPU because I noticed the best temp drops on my CPU after undervolting the 1070. Hope it's a faulty Bios or OC software.
As the coffeelake processor supports higher clock rate than kabylake or older processors, you may easily notice that the Alienware 13R3, 15R3, 15R4, 17R4, 17R5 generates more heat compared to previous generation platforms. Therefore Dell engineering optimized the system to tolerate higher temperature from design perspective, and the new specification for Alienware 15 R4 / 17 R5 is: CPU peak temperature will be no greater than 100 degree C.
Troubleshooting steps for CPU high temperature:
1. Update the BIOS/driver to up-to-date from Dell support site.
2-1. Open “ Alienware Command Center”
2-2. Configure the Active System Theme - Thermal as “ Performance mode”
2-3. Configure the Active System theme - Overclock to “ Overclock O2 “ (if applicable)
Also, please run the HWINFO app and send us a picture of the temps so we can check the averages.
Dell | Alienware Technical Support
Opening your case does not void your warranty! Check out Alienware Artificial Intelligence
I can't always reply to PM's, if you need extra assistance visit:
There are thinner, lighter laptops from other manufacturers running the same spec at much lower temps so please stop lying to us and yourselves.
Ive come a across a newer MUCH MORE annoying problem now. Everytime i connect a bluetooth controller to my AW15, Wifi speeds drop from 20mbps to 0.25mbps, as soon as i disconnect my controller speeds shoot back up again. THIS IS BY FAR THE WORST THING IVE PUT MY MONEY INTO.
I've attached a screenshot from HWMonitor as per your request. KEEP IN MIND : THESE ARE TEMPS AFTER UNDERVOLTING BY QUITE A MARGIN.
@CJORACLE I suggest you undervolt the cpu and lower the multiplier to 36x using throttlestop - This should cap the boost clock at 3.58ghz
And then use MSI Afterburner to get yourself a more agressive and steeper voltage frequency curve.
This ofcourse requires a lot of trial and error and is a risk if done without care. But hey it increases the life span of the internals and I don't need the CPU to boost upto 4.1 for most games I play, so it's a downgrade I'm willing to make.
@Kaushikhcould u share the picture of ur frequency curve on msi afterburner, also how much undervolting u did on your cpu.
Personally for me undervolting cpu did nothing to the temps. I was thinking of undervolting Gpu as well, as GPU stays pretty cool at 70c under heavy gaming
Your suggestion worked like a charm.
Thanks a ton!!! After decreasing multiplier in throttlestop to 35 and undervolting to good -.140.. i i w able to get the temps to 75 max during farcry 5 on ultra, that to without any fps loss.
Earlier the temps were 90 on an average in jus 20 mins of gameplay.
This is by far the best possible solution i have found if u one is not ready to repaste thermals on the new laptop.@
@CJORACLE Hey man that's good news, I'm glad I could help. We seem to have the exact same thermals with and without the undervolts and underclocks so its very unlikely for us to have both recieved duds. Seems to be a bios problem or poor quality thermal paste. Also, I suggest you set up a couple different profiles in throttlestop. I run stock multipliers for rendering which doesn't use the GPU at all. 36x for open world games and a lower 32x for sports and racing titles.
I also got a AW15 R4 with the 8750H and Nvidia 1070 GPU.
Just want to add some additional information to the thread.
I've been able to undervolt the CPU core with -160 and CPU cache -125 now without touching the multiplier at all.
BD PROCHOT is unchecked and SpeedStep is checked.
Speed Shift - EPP is set to 0.
The system has been stress tested for about 1h with no BSOD or any other error. Temp is around 75-80c while playing world of warcraft and browsing the web at the same time.
Alright. So I've been running this for a good 2 days and system just crashed with a BSOD related to CPU.
STACK_COMMAND: kb FOLLOWUP_NAME: MachineOwner MODULE_NAME: GenuineIntel IMAGE_NAME: GenuineIntel DEBUG_FLR_IMAGE_TIMESTAMP: 0 FAILURE_BUCKET_ID: X64_0x124_GenuineIntel_PROCESSOR_MAE BUCKET_ID: X64_0x124_GenuineIntel_PROCESSOR_MAE Followup: MachineOwner
Changed the CPU core voltage to -155 instead of -160.
I wanted to take Kaushikh's advice and reduce the multiplier to 36X, but I'm not seeing it reflected in the FID. I'm using Throttlestop 860. I checked the Set Multiplier box, set it to 36T, and hit Save, but no apparent effect. Are there any other boxes (I have SpeedStep, BD PROCHOT, C1E, and Task Bar checked now) that need to be checked?