PowerEdge: BlueField-3 DPU PCIe Initialization Failure
Summary: Encountering PR8 errors in the LifeCycle (LC) log when using a BlueField-3 (BF3) Data Processing Units (DPU) card (DPN: HFWRM).
Symptoms
The LifeCycle controller logs (LC log) report the following errors due to repeated Peripheral Component Interconnect Express (PCIe) initialization failures:
2025-07-27 17:38:59 294 PR8 Device not detected: Nvidia Network Adapter - 5C:25:73:5A:4C:B8(NIC in Slot 33 Port 1 Partition 1)
2025-07-27 17:38:58 293 PR8 Device not detected: Nvidia Network Adapter - 5C:25:73:5A:4C:B9(NIC in Slot 33 Port 2 Partition 1)
2025-07-17 17:30:57 189 PR8 Device not detected: Nvidia Network Adapter - 5C:25:73:5A:4C:B8(NIC in Slot 33 Port 1 Partition 1)
2025-07-17 17:30:57 188 PR8 Device not detected: Nvidia Network Adapter - 5C:25:73:5A:4C:B9(NIC in Slot 33 Port 2 Partition 1)
2025-05-11 17:29:46 46 PR8 Device not detected: Nvidia Network Adapter - 5C:25:73:5A:4C:B8(NIC in Slot 33 Port 1 Partition 1)
2025-05-11 17:29:46 45 PR8 Device not detected: Nvidia Network Adapter - 5C:25:73:5A:4C:B9(NIC in Slot 33 Port 2 Partition 1)
Cause
The issue has been confirmed and was resolved in firmware version v32.46.3048, released on 14-Aug-2025.
This is a known NVIDIA issue. The device firmware impacted the training process during PCIe network initialization, and the matter has been resolved starting from firmware version v32.46.3048.
The firmware algorithm has been optimized to improve PCIe link stability. It is confirmed that newly shipped NVIDIA cards already include the updated firmware version v32.46.3048 or later.
Resolution
Do NOT dispatch a replacement network card immediately. First, run a full power cycle to verify whether the card becomes detectable again.
- If the failed card recovers after the power cycle:
Update the BlueField‑3 firmware to v32.46.3048 or later, and ensure that the BFB image is aligned with the updated firmware. You can download the appropriate files using the NVIDIA DOCA Software Framework:
https://developer.nvidia.com/doca-downloads?deployment_platform=BlueField&deployment_package=BF-FW-Bundle&installer_type=BFB - If the card remains undetectable after the power cycle:
Proceed with dispatching a replacement card to resolve the issue.