Hello everyone, My device suddenly shut down completely without rebooting. After checking the logs, I found these errors: plaintext cx7-pcie-hotplug MTKP0001:00: Cable removal pcieport 0000:00:00.0: AER: Correctable error message received pcieport 0002:00:00.0: PCIe Bus Error: Physical Layer, RxE…

[image] yousimu.wang: cx7-pcie-hotplug Do you have anything connected to 400Gb/s port at the back?

I’m quite sure my Spark is randomly powering off because of overheating. I’ve been monitoring its temperature curve, and it’s clear it’s just not built for long-running agent tasks. I’m really frustrated with this. I’m wondering if anyone else on the forum has encountered the same issue.

[image] DGXSPARK temperature too high, automatic shutdown。 DGX Spark / GB10 Room temperature is 18 degrees, DGXSPARK running COMFYUI to generate video (about 10 minutes), often turns off automatically due to overheating. Now I dare not use COMFYUI, how to set t…

The kern.log shows a ConnectX-7 hotplug event immediately followed by PCIe Physical Layer errors on the GPU root ports: 18:03:35.952530: cx7-pcie-hotplug MTKP0001:00: Cable removal 18:03:35.952551: pcieport 0002:00:00.0: PCIe Bus Error: severity=Correctable, type=Physical Layer, (Receiver ID) 18:03…

Your suggestions are highly valuable and well worth trying out. yousimu@spark-02df:~$ sudo dmesg | grep -iE "aer|pcie|mlx5" [ 0.000000] Kernel command line: BOOT_IMAGE=/boot/vmlinuz-6.17.0-1014-nvidia root=UUID=c5b52ea7-8a82-4154-853a-ef22b1878d4c ro init_on_alloc=0 iommu.passthrough=0 console=…

I will try various cooling methods for this device to fix the random unexpected shutdown issue. The machine shut down several times automatically these days when the temperature was extremely high, so I dared not keep it running. Recently the weather has turned cool and no more shutdowns occurred. …

I would probably set the mac mini somewhere else if your dealing with high temps since the mini puts out heat of its own.

Looking at your boot dmesg from spark-02df. CX7 enumerates correctly at 1.2–3.6s — all four ports report firmware 28.45.4028 and full 126.028 Gb/s available PCIe bandwidth (32.0 GT/s PCIe x4 link). The issue starts after the hotplug driver initializes at 8.0s. Immediately after cx7-pcie-hotplug: H…

any results? for the same workload, does the temp decreased?

Dgx spark shut down without rebooting

Accelerated Computing DGX Spark / GB10 User Forum DGX Spark / GB10

PILCOTHINK May 12, 2026, 7:31am 8

There are quite a few people here with similar experiences.

Topic		Replies	Views
DGXSPARK temperature too high, automatic shutdown。 DGX Spark / GB10	205	11429	October 17, 2026
DGX Spark fans stop when the screen goes dark or running from SSH — box gets too hot to touch* FIRE HAZARD DGX Spark / GB10 dgx-spark-issue	18	1804	September 23, 2026
Spark abruptly shuts down DGX Spark / GB10 dgx-spark-issue	47	3577	October 9, 2026
DGX Spark constantly shut down after first set up DGX Spark / GB10	8	735	January 11, 2026
DGX Spark (GB10) reproducibly hard powers-off under GPU load — fully updated, zero crash capture DGX Spark / GB10 boot , kernel , ota , dgx-spark-issue	13	1385	June 14, 2026
DGX Spark reboots every ~20 minutes DGX Spark / GB10	11	771	September 3, 2026
DGX Spark. low fan speed, high temps, device very hot DGX Spark / GB10 kernel , gpu , fan-facts , debugging-and-troubleshooting	61	6990	August 3, 2026
DGX Spark GB10: partnerdiag PowerStress reproducibly hard-powers-off the box — acpitz 88→97.8°C in 5s, all other field tests pass DGX Spark / GB10 Projects dgx-spark-issue	9	797	July 24, 2026
DGX Spark stability / out of RAM / overheating DGX Spark / GB10 llama , dgx-spark-issue	51	4612	September 7, 2026
DGX Spark Performance Degradation - GPU Power Draw Issue DGX Spark / GB10 power , performance , llama , dgx-spark-issue	69	6193	June 15, 2026

Dgx spark shut down without rebooting

Related topics