Xid
A system of error codes written to the kernel log when the NVIDIA GPU driver detects hardware or firmware anomalies. Each number corresponds to a specific fault root cause (for example, Xid 79 indicates the GPU has fallen off the PCIe bus) and is used for operations triage.
Timeline (last 90 days)
Xid is the standard error code system reported by the NVIDIA GPU driver, with different codes corresponding to hardware or driver problems of different natures.
Xid 13/31 is mostly related to illegal memory access or out-of-bounds access by user applications, Xid 48 usually indicates a double-bit ECC error, Xid 79 indicates the GPU has fallen off the bus, and Xid 62/64 is often related to page table or internal microcontroller anomalies.
Some Xid errors are triggered by application bugs and can be recovered by restarting the process or node, while others foreshadow hardware degradation of memory or power supply, and repeated auto-remediation would instead mask the problem.
Users can individually decide whether to enable auto-remediation for each type of fault and each Xid error code.
Xid messages are written to the system log by the NVIDIA driver when it detects a GPU anomaly, typically appearing in dmesg or /var/log/syslog on Linux, in a format like "NVRM: Xid (PCI:0000:3b:00): 79, pid=12345".
AI-Node-Sentinel's Xid lookup table covers only seven codes (13, 31, 48, 74, 79, 119, 120), and all codes outside the table return None and are escalated to a human, with refusal to guess being the most important safety property of the pipeline.
NVIDIA's Xid numbering system covers GPU hardware and firmware errors, with each number corresponding to a specific class of failure root cause, for example Xid 79 indicates the GPU has fallen off the PCIe bus, Xid 48 corresponds to an uncorrectable double-bit ECC memory error, Xid 74 involves an NVLink link failure, and Xid 119/120 relate to GSP firmware timeouts.
All Facts (7)
Xid is the standard error code system reported by the NVIDIA GPU driver, with different codes corresponding to hardware or driver problems of different natures.
50%UnverifiedXid 13/31 is mostly related to illegal memory access or out-of-bounds access by user applications, Xid 48 usually indicates a double-bit ECC error, Xid 79 indicates the GPU has fallen off the bus, and Xid 62/64 is often related to page table or internal microcontroller anomalies.
50%UnverifiedSome Xid errors are triggered by application bugs and can be recovered by restarting the process or node, while others foreshadow hardware degradation of memory or power supply, and repeated auto-remediation would instead mask the problem.
50%UnverifiedUsers can individually decide whether to enable auto-remediation for each type of fault and each Xid error code.
50%UnverifiedXid messages are written to the system log by the NVIDIA driver when it detects a GPU anomaly, typically appearing in dmesg or /var/log/syslog on Linux, in a format like "NVRM: Xid (PCI:0000:3b:00): 79, pid=12345".
50%UnverifiedAI-Node-Sentinel's Xid lookup table covers only seven codes (13, 31, 48, 74, 79, 119, 120), and all codes outside the table return None and are escalated to a human, with refusal to guess being the most important safety property of the pipeline.
50%UnverifiedNVIDIA's Xid numbering system covers GPU hardware and firmware errors, with each number corresponding to a specific class of failure root cause, for example Xid 79 indicates the GPU has fallen off the PCIe bus, Xid 48 corresponds to an uncorrectable double-bit ECC memory error, Xid 74 involves an NVLink link failure, and Xid 119/120 relate to GSP firmware timeouts.
50%