We have 7490's and about a third of them run at 100% for NO REASON.
I LITERALLY JUST UNBOXED 5 OF THEM AND 2 OUT OF 5 ARE BLOWING AT 100% FAN POWER.
The other 3 are silent. This is straight out of the box behavior, with stock power adapters.
Disabling Intel SpeedStep in the BIOS fixes this temporarily. I rolled them up to 1.3.3 (latest bios) and saw no change. The laptops are actually hot. Like this is clearly a thermo problem if 3 of them are fine out of the box.
Same issue here. Unboxed and fan running constantly, the machine gets quite hot from the bottom. About to upgrade BIOS to 1.3.3, but according to the "what's new" I don't expect any fixes on this. Did youget any good advice regarding the issue?
My Latitude 7490 had the same symptoms, fan running all the time and the system was hot. In the Task Manager I could see that the WSAPPX process was taking much of the CPU time, at minimum 25%. Via Google I found instructions to change in Local Group Policy Editor setting both for Computer Configuration and User Configuration:
> Administrative Templates > Windows Components > Store > Enable the Store application
These were originally set to "Not configured", so it seemed ambivalent. Setting to Enable made the WSAPPX to behave, running only immediately after start up to check the updates for apps installed via the store. The original instructions said to disable the setting, but then you don't have the Store, so this no option.
But my 7490 is still having major problems with the 1803 update, still waiting for MS to come up with a proper update... =(
Be mindful that you need to give Windows 10 time to settle after setup (it can take hours before it's recompiled the libraries or whatever it's doing, especially if you start applying full load immediately). Same thing happens with an iPhone immediately after an OS upgrade btw, that's why people are always complaining about reduced battery performance at each and every new version.
Note: There is definitely other stuff that can cause the CPU to get stuck in high gear. We've found the audio driver to be a regular suspect, upgrading/downgrading or just remove and reinstall usually fixes it.
The only way I was able to get rid of this (partially) is using the software called ThrottleStop.
The CPU model I have is the i7-8650u.
What I did is : - Undervolting as much as possible (still with a rock stable system) cpu core = -110.4mV cpu cache = -110.4mV intel gpu = -60.5mV system agent = -60.5mV - Reducing the package power limits (TDP) From 22 / 44 to 15 / 20 (watts)
This allow to reduce the temperature without impacting the performance
I have similar issues, for instance right after booting a fresh win 10 install first time, some windows processes (windows update, windows security essential and alike) keep the CPU at 100% and then I get BSOD after 1-2 minutes ( IRQL_NOT_LESS_OR_EQUAL, BAD_POOL_HEADER) and the crashing module is almost always the ntoskrnl itself, so it does not point me in any direction). It does not always do this, only the first few boots and then things calm down a bit.
I got the motherboard replaced and it's still the same. Tried with another pair of RAM modules as well. Same result.
I reinstalled win10 3 times, played with settings of speedstep, speedshift, turbo and others, no luck yet.
I'm on latest BIOS.
Same issue with default drivers and newest drivers as well. Dell diagnostics do not give any error.
During torture tests or during very high load, sometimes, but interestingly not always the CPU temperature goes as high as 95°C and the CPU fan is at 6500RPM. Ambient temperature is 23-24°C, air conditioned.
Strangely I got BSOD every time I try to start PassMark BurnInTest, I also get BSODs frequently when I start AIDA64. Even when CPU is below 70°C, so BSODs may not be related to heating issues at all.
After playing with the settings and reinstalling windows about 15 times, I can say that the BSOD is somewhat related to Driver Operation / RAID On mode (in some cases, the failing module was storport.sys, most of the times it was ntoskrnl.exe). When enabled, even there is no raid storage, only a single, no raid one, it crashes randomly, but consistently under high load. Using AHCI, I haven't experienced any BSOD. I'm not sure that going with AHCI mode solves the issue or just lowers the chance of the crashes, I keep testing.
About the overheating I got quite confusing results during my testing. Sometimes right after first boot it just starts the thing, fan @ 6500RPM, 99°C, constant max turbo frequency (7300U@3.5GHz), even when there is only 60% CPU load or less (!), that's very strange to me. And it settles down very slowly if at all.. suspicious processes are wsappx, tiworker, windows update, onedrive (even when it's disabled during install, it starts with very high cpu load). I see a lot of system interrupts as well, sometimes up to 8% cpu load.
I experienced that when running on battery, almost there isn't any heating issue. Of course, the performance is lower in this case.
But when plugged in, the power management seems to be a lot more aggressive so it gives crazy temperatures and fan speeds under load. What's strange to me that even at 99°C it stays at max turbo frequency and then starts heavy thermal throttling, instead of going down to a lower frequency and VID to protect the CPU. I find that thermal design or power management is very poor with this model.
DELL-Alasdair R
6 Operator
•
754 Posts
115498
0
Posted May 2nd, 2018 04:00
Hi dfriedmango,
I sent you a PM...