GPU_NVLINK_Link_training_Failure_Event : LTSSM_FAULT
LTSSM_FAULT
A physical layer fault, which is a disruption or fault at the physical layer, will lead to an LTSSM FAULT. In this fault scenario, error occurs on the GPU side of the NVLINK.
Severity
Error
Alert Category
Interconnect/Link Error
User Action
- Check if errors persist across resets.
- Reseat the node.
- If the issue persists, collect a diagnostic bundle using NVDebug.
- Contact Lenovo Support.NoteSome events are informational and require no action.
Give documentation feedback