Hello, I have a problem with my Precision 5510 laptop. After adding a second NVMe SSD + fresh Win10 reinstall, system overheats after ~5-10min of watching Youtube 4K60 video, CPU clock drops to 800MHz, laptop becomes super slow. After some time with 0% load clock frequency recovers. All drivers installed from Dell website, as well as Power Manager.
System Configuration: i7 6820HQ, Quadro M1000M, 2x8GB, Samsung 850 EVO 500GB SATA + Samsung 970 PRO 512GB NVMe (System drive) + Windows 10 PRO + 2 external monitors via Pluggable Thunderbolt 3-> 2x DP adapter (Dell UP2516D + Dell U3014).
Laptop is free of dust, CPU temperatures are normal, however, I still removed the heatsink and replaced the thermal compound to Arctic MX-4. That did not help: fans remain loud while CPU+GPU temps are low. Something else overheats and CPU clock drops to 800MHz. As others have suggested in other posts in this forum, I tried unplugging internal battery, resetting the laptop by holding power button 30sec. Original 130W PSU seems to be discovered properly in the BIOS. Bios is not reporting any events. Latest BIOS is installed, reinstalled. Undervolting the CPU with Intel XTU does not help significantly.
I now opened the bottom cover, placed an external fan over the chip near the battery (+additional heatsink over hot VRMs) and I no longer get that issue - system remains stable. However, I cannot use the laptop that way. Is this some known issue with this laptop? Maybe some software could fix this? Laptop is technically still under warranty, however, totally unusable in current configuration.
Update: after further testing I realized that those are VRMs between CPU and GPU overheating and causing the problem. I now placed an external fan over them instead of the bottom chip and system remains stable. I still see the Power Limit Throttling (CPU clock go to 800MHz), but only with synthetic 100% CPU+GPU load which is unrealistic.
After further googling I found that many others experience same/similar issues with both XPS 9560 and even 9570 models which is troubling. I am not a big fan of giving my laptop away for days for warranty service. How likely is it that this problem would be fixed while still maintaining good laptop performance? Or is it a good time to get back to thick laptops (never had any issues with both Latitude E6540, E6440 + i7-4800MQ that I owned previously and that benchmarked similarly regarding CPU performance).
P.S. I am also attaching a photo of sensor data when watching Youtube video (4K60). Even with near 0 load the fans spin at annoying ~4000rpm:
What is the first Ambient temperature under Dell EC in HWinfo64? Should be the sensor near the VRMs, and should trigger throttling.
If this is a new condition, try reinstalling Intel DPTF which supposedly does the throttling, maybe other chipset drivers too.
It is well known that these VRMs deserve some cooling. Many users stack a few Arctic thermal pads to send a bit of heat to the backplate (but don't overdo, makes things worse). You can look up iunlock's mod at NBR forums for a more proper solution.
But simple video playback shouldn't be causing this, unless it really loads up the GPU. Try reinstalling Intel and NVidia drivers too, using Display Driver Uninstaller.
Thank you for your response! The first DELL EC Ambient temperature is reading 46*C with web-browsing load, fans at 4100rpm, actual ambient temperature of ~22C. When giving some load (youtube+Cinebench) I get inconsistent results: sometimes the CPU would PL throttle to 2.5GHz when first Ambient temperature sensor reads~65C, sometimes it would hold the load up to over 80C and then CPU clock would instantly drop to 800MHz, afterwards slowly recover to normal. Hopefully, tomorrow I will get another SSD so I could both, eliminate SSD from the equation and also install a fresh copy of Windows.
Also, I was thinking of adding some quality Thermal Grizzly Minus Pad 8 pads in that area (over VRMs, as suggested in notebookreview forum thread about XPS 9560 CPU + GPU Temperatures), but I don't see them helping to a sufficient degree. I believe that cooling fans shouldn't be running at these speeds with ~0 load and Dell Power Manager set to Optimized.
Okay, so I now installed a SATA HDD to the laptop (also leaving the suspicious Samsung 970 PRO NVMe SSD installed as secondary drive with my throttling OS), along with a fresh copy of Windows and the problem seems to have magically disappeared. Fans run at slightly more reasonable speeds, frequency no longer drops to 800MHz under normal use under load. I now only get to see Power Limit Throttling when using AIDA64 CPU+GPU stress test, which I think is reasonable (somebody, please confirm).
The problem was most likely software-related, however, it is bizarre that a cooling fan placed over VRMs somewhat solves it. I now tried some simple solutions like reinstalling drivers (including Intel DPTF), Dell utilities (including Power Manager), uninstalling Intel XTU on my old throttling OS, however, that did not help. It will probably be easier to reinstall Windows on that main NVMe drive, unless anyone else has an understanding of what might cause such problem. I will post an update after I do that to confirm that the issue was solved.
Check and make sure your BIOS SATA setting isn't RAID. I had increasingly poor performance, overheating, etc... and I was seeing the same reduction to 800 MHz until I changed the SATA mode to AHCI. It's now like a new machine. (Dell apparently used RAID mode because it was the only way to have NVMe with Win7 - but it is non-optimal with Win10).
To change, type msconfig in the search bar, and run as administrator. Select the 2nd tab, and select safe boot minimal.
restart, and immediately hit F12 to go get the startup menu, then go to BIOS Setup. Change from RAID to AHCI.
then let it boot into safe mode.
Then type msconfig again, run as administrator, and change back to normal boot.
Reboot to Windows, then run Dell's update utility to install the SATA AHCI drivers.
This was an instant, night and day difference on my Precision 5510.
i totally agree with @Sh4veD4ve i was in this nightmare as my temp was all below 70c and still my cpu throttle down to 800mhz instantly when i open a game or put any load on my processor, 3 months with re pasting my cpu+gpu trying drivers uninstalling some from device manager, nothing seems to solve my problem till i changed my sata configuration from raid to AHCI MODE and all my problems are gone!
hope every one who has the same problem with xps or precision to see this comment and found this fix!
thank you sh4ved4ve for bings this up!