*OFFICIAL* Alienware m15 Owner's Lounge

I used Kryonaut the first time around. It lasted for a few months tops. The reality is that this model has a sink fit issue and needs a thick paste to compensate.

 

That's why I tried switching to MX-4. I am looking at a 20C difference between cores so I am wondering if there is a heatsink issue.

 

Well, again, you can try Phobya and repadding, given the reasonable rationale. The pads supplied by Dell are very hard so surely they reduce sink pressure to some degree, but don't expect miracles. Max temp 97C and core temp differenc reduced to 10C. Some people have had great success with LM too, although I'm personally skeptical, given the suspected sink pressure issue.

 

Tried searching the forums but to no avail. Does anyone have an m15 with OLED screen? I do and HDR doesn't work correctly it seems. When I turn HDR on in windows settings the first thing to notice is windows desktop becomes dim and bland looking. The next thing is netflix HDR shows/movies looks as dull and bland as windows desktop. I also tried shadows of the tomb raider and noticed the same. The other issue is fps is limited to around 40 fps.

Any help is appreciated.

 

So this is weird and I feel like its gonna sound like I am making a mistake in my typing. So I have a 2070 Max-Q in my m15. Disclaimer : Not about mining, but about the 2070 Max-Q.

- The stock curve maxes out at 1920core and 1100mV and has a reported wattage of 70W in Phoenixminer.
- Reducing the core clock and mem clock by -500, makes Phoenixminer show a 90W power draw.
- I can make it a straight voltage line with 700mV @ 2000 core clock, and Phoenixminer shows 66W.

Is there any reason this can make sense? The GPU temps have proven that the voltages are what I typed above, they go up/down when wattage goes up/down as reported by Phoenixminer. Why would the stock curve have a 70W power draw with 1100mV but reducing the core clock and mem clock by -500 (which drops the whole curve A Lot), Increases the wattage A Lot while the voltage stays the same?

 

My impression of HDR support in Windows, based on limited experience, is that it's a sad joke, however, I have also seen the issue you describe. The "solution" was to play around with colour output settings in Nvidia control panel (IIRC RGBA 32bit worked best) - that was with an external monitor though, which had to be put in HDR mode.

 

delete

 

Power is voltage x current. Although you lower the voltage, the core clock at that reduced voltage may increase if you are neither power limit throttling or thermal throttling. Increasing clocks translate into increasing power draw. Generally speaking, when I try to undervolt my GPU, my goal is to prevent power limit throttling. I keep adjusting the curve until I crash at which point I raise the voltage just slightly.

 

I am confused. Are you saying its part of that NV 'boost clock' tech where they overclock/underclock depending on temps/voltages? I would think the wattage should be the same, if not less, if core clock is minus 500 and mem clock is minus 500. I can see how the other 2 scenarios make sense, I just don't see why the GPU wants to use +20W from reducing core and mem clocks by 500.

 

So, I've been having these issues where on boot fans would sometimes go to 100% and ePSA would kick off and say something like "Support Assist detected an issue with system fan". I must have gone through 20 support emails, they released some BIOS updates, to no avail. Clearly thre is no actual hardware issue.

Uninstalling Support Assist solved the problem....