ESXi host PSOD with LINT1/NMI when powering on a VM with NVIDIA GPU Passthrough

Products

• VMware vSphere ESXi

Issue/Introduction

• An ESXi host crashes with a Purple Screen of Death (PSOD) when powering on a Virtual Machine (VM) configured with an NVIDIA RTX Pro 6000 GPU in PCI Passthrough (DirectPath I/O) mode.
• The PSOD screen and vmkernel logs display the following backtrace: 
     PSOD Message: @BlueScreen: LINT1/NMI (motherboard nonmaskable interrupt), ApeiHESTNmiHandler has diagnosed a fatal error.
   Backtrace for Current CPU - X:
     0x452b80002d40:[0x42001bd7c480]PanicvPanicInt@vmkernel#nover+0x20c stack: 0xfbbc1d995e954, 0x42001bd7c480, 0x0, 0x420000000001, 0x42001bd7c480
     0x452b80002df0:[0x42001bd7cbfc]Panic_NoSave@vmkernel#nover+0x4d stack: 0x452b80002e50, 0x452b80002e10, 0x452b80002eaf, 0x452b80002e8f, 0x0
     0x452b80002e50:[0x42001bd78b23]NMI_Interrupt@vmkernel#nover+0x4d8 stack: 0xa6c6865aaa86865e, 0x9df614fa91b614fe, 0x9efea3d892bea3dc, 0xa5ce3178a98e317c, 0x18638d7714238d73
     0x452b80002f10:[0x42001c2a7404]IDTNMIWork@vmkernel#nover+0x95 stack: 0x0, 0x42001c2a886d, 0x0, 0x42001c2a20c7, 0x103
     0x452b80002f30:[0x42001c2a886c]Int2_NMI@vmkernel#nover+0x9 stack: 0x103, 0x0, 0x0, 0x1, 0x3d4f
     0x452b80002f40:[0x42001c2a20c6]gate_entry@vmkernel#nover+0xa7 stack: 0x0, 0x1, 0x3d4f, 0x430485001830, 0x1
     0x453c0349ba38:[0x42001bc90ed7]Power_ArchPerformWait@vmkernel#nover+0xab stack: 0x420040001880, 0x0, 0x0, 0x420040000000, 0x420040000000
     0x453c0349ba40:[0x42001bc91051]Power_ArchSetCState@vmkernel#nover+0xba stack: 0x0, 0x0, 0x420040000000, 0x420040000000, 0x0
     0x453c0349ba90:[0x42001c2db75d]CpuSchedIdleLoopInt@vmkernel#nover+0x292 stack: 0x0, 0x7fffffffffffffff, 0x1, 0x7fffffffffffffff, 0x453c0da1f100
     0x453c0349bb00:[0x42001c2dfc88]CpuSchedDispatch@vmkernel#nover+0x1e31 stack: 0x1, 0x420040001040, 0x420040001110, 0x420040001128, 0x420040001040
     0x453c0349bd40:[0x42001c2e069a]CpuSchedWait@vmkernel#nover+0x35b stack: 0x9d, 0x0, 0x1000000000000, 0x9d, 0x8000000
     0x453c0349beb0:[0x42001bda0256]vmk_WorldWait@vmkernel#nover+0x5f stack: 0x10749d08d02, 0x453c0349bf50, 0x45bd18dce348, 0x42001d8dd41b, 0x4331a3c79080
     0x453c0349bee0:[0x42001d8dd41a][e-mail]#0.0.0.1+0xbbf stack: 0x10749d144a4, 0x4331a3c79158, 0x10749d061f3, 0x800000000, 0x0
     0x453c0349bfa0:[0x42001bda00fc]vmkWorldFunc@vmkernel#nover+0x31 stack: 0x42001bda00f8, 0x0, 0x453c0349f000, 0x453bccc9f100, 0x453c0349f100
     0x453c0349bfe0:[0x42001c2dddfe]CpuSched_StartWorld@vmkernel#nover+0xbf stack: 0x0, 0x42001bd455b0, 0x0, 0x0, 0x0
     0x453c0349c000:[0x42001bd455af]Debug_IsInitialized@vmkernel#nover+0xc stack: 0x0, 0x0, 0x0, 0x0, 0x0
   LIN1 related errors in the logs
     LINT1/NMI (motherboard nonmaskable interrupt), ApeiHESTNmiHandler has diagnosed a fatal error.
     2026-06-11T06:02:14.018Z cpuX:2099477)@BlueScreen: LINT1/NMI (motherboard nonmaskable interrupt), ApeiHESTNmiHandler has diagnosed a fatal error.

Environment

vSphere ESXi 8.x

Cause

The cause of the PSOD is due to a fatal hardware-level PCIe error reported by the PCIe port hosting the NVIDIA GPU. The crash is triggered during device initialization when PCIe Hot-Plug capabilities or bridge-based device reset routines are actively executed. Because Broadcom does not support PCIe Hot-Plug operations for passthrough GPU devices, this creates a critical incompatibility where either the device firmware, the guest operating system driver, or the motherboard BIOS fails to properly arbitrate the Hot-Plug signaling. This structural breakdown leaves the PCIe bus in an unresponsive state during the virtual machine's power-on sequence, which the motherboard firmware subsequently intercepts and flags as a fatal Non-Maskable Interrupt (NMI).

Resolution

To workaround this issue follow the below steps:

NOTE: Setting enablePCIEHotplug=FALSE prevents ESXi from enabling hot-plug during server boot, even if the hardware supports it.

 

1. Disable PCIe Hot-Plug by running the following command on the ESXi host:
   
   esxcli system settings kernel set -s "enablePCIEHotplug" -v "FALSE"
   
   

2. Modify the passthrough configuration

   • Backup the file
     cp /etc/vmware/passthru.map /etc/vmware/passthru.map.bak
     
     
   • Edit the file using vi/etc/vmware/passthru.map.
   • Locate the line:
     
     # NVIDIA (FLR issue on Ampere and Hopper GPUs)

     10de ffff bridge false
     
     
   • Change it to:
      

     # NVIDIA (FLR issue on Ampere and Hopper GPUs)

     10de ffff default false
     
     
3. Reboot the ESXi host to apply the changes.

4. Verify that PCIe device hot-plug is disabled by entering the command:

   esxcli system settings kernel list -o enablePCIEHotplug
   
   The entry, "FALSE," should be displayed under the Configured column:
   
   Name                Type     Configured  Runtime Default  Description 
------------        ----     ---------   ------  -------  -----------
enablePCIEHotplug   Bool      FALSE      FALSE    TRUE   Enable PCI-E Native Hotplug support 

Additional Information

VM with an NVIDIA GPU in passthrough mode fails to power on with the error: "Module 'DevicePowerOn' power on failed".

VM poweron fails due to hotplug bus reset after adding a third NVIDIA passthru device