DCGM_FI_DEV_ROW_REMAP_FAILURE pending row remap failed bank row reserve exhausted

HBM row remapping failed: the GPU has exhausted its reserve of spare memory rows and can no longer retire failing ones. This is a terminal hardware state, the GPU will keep producing uncorrectable ECC errors and must be replaced. This entry explains how to confirm the cause, apply the fix, and separate it from adjacent telemetry failures.

Symptom

DCGM_FI_DEV_ROW_REMAP_FAILURE pending row remap failed bank row reserve exhausted

Root cause

HBM row remapping failed: the GPU has exhausted its reserve of spare memory rows and can no longer retire failing ones. This is a terminal hardware state, the GPU will keep producing uncorrectable ECC errors and must be replaced. The decisive evidence is the first log line that precedes "DCGM_FI_DEV_ROW_REMAP_FAILURE pending row remap failed bank row reserve exhausted" and differs from a healthy run.

Recommended fix

drain the GPU and stop scheduling work on it. Resume the job from a checkpoint written BEFORE the first remap failure, results computed after it are suspect.

How Denpex helps

Denpex investigates DCGM_FI_DEV_ROW_REMAP_FAILURE pending row remap failed bank row reserve exhausted using the evidence you provide or your connected workload collects. Earlier rank, host or application evidence is needed to distinguish an initiating failure from a downstream report.

The literal signature is "DCGM_FI_DEV_ROW_REMAP_FAILURE pending row remap failed bank row reserve exhausted". It is a hardware failure associated with NVIDIA telemetry, DCGM, NVML, power, and cooling. The line identifies the failing operation or subsystem, while the surrounding evidence decides whether it is the initiating fault or a downstream symptom.

HBM row remapping failed: the GPU has exhausted its reserve of spare memory rows and can no longer retire failing ones. This is a terminal hardware state, the GPU will keep producing uncorrectable ECC errors and must be replaced. The failure becomes visible at this call site because the operation first requires the missing resource, valid state, healthy peer, or correct result. Earlier log lines and a known-good control carry more causal value than the final wrapper exception.

• The workload stops or loses forward progress after emitting "DCGM_FI_DEV_ROW_REMAP_FAILURE pending row remap failed bank row reserve exhausted".
• A retry on the same configuration reproduces the failure because the causal state has not changed.
• The outer framework exception can hide the rank, node, allocation, or dependency that failed first.
• Increasing timeouts or reducing workload size can suppress the symptom without correcting the cause.

• NVIDIA telemetry, DCGM, NVML, power, and cooling
• production-shaped multi-accelerator workloads
• containerized and bare-metal deployments of the same stack

Use this checklist to test the hypothesis against a small reproduction. No single line proves the root cause, so preserve the preceding events and compare one variable at a time.

• ✓Find the first occurrence of "DCGM_FI_DEV_ROW_REMAP_FAILURE pending row remap failed bank row reserve exhausted" and preserve at least 100 lines before it.
• ✓Identify which rank, node, device, or process emitted the first related warning.
• ✓nvidia-smi -q -d ROW_REMAPPER (check "Remapping Failure Occurred" and the pending/available counts); nvidia-smi -q -d ECC for the volatile/aggregate DBE trend; dcgmi diag -r 3.
• ✓Repeat the same input after the targeted change and require the signature to disappear.
• ✓Resume from checkpoint predating the first remap failure only after the control passes.

• HBM row remapping failed: the GPU has exhausted its reserve of spare memory rows and can no longer retire failing ones. This is a terminal hardware state, the GPU will keep producing uncorrectable ECC errors and must be replaced.
• The decisive evidence is the first log line that precedes "DCGM_FI_DEV_ROW_REMAP_FAILURE pending row remap failed bank row reserve exhausted" and differs from a healthy run.
• A nearby failure remains a competing hypothesis until a control separates configuration, capacity, transport, and hardware causes.

DCGM_FI_DEV_ROW_REMAP_FAILURE pending row remap failed bank row reserve exhausted

Use this text as a lookup key in logs and upstream issue trackers. It is not presented as a captured customer log. Confirm the cause from your own preceding events, versions, configuration and the cited references.

drain the GPU and stop scheduling work on it. Resume the job from a checkpoint written BEFORE the first remap failure, results computed after it are suspect. This changes the condition in the causal diagnosis instead of hiding the outer exception. The repeated control proves ownership before recovery from checkpoint predating the first remap failure.

# Preserve evidence before restarting
dmesg -T | grep -iE 'NVRM|Xid|AER|NVLink|ECC'
nvidia-smi -q
dcgmi diag -r 3

# Find the exact signature in the complete log
rg -n -F -- "DCGM_FI_DEV_ROW_REMAP_FAILURE pending row remap failed bank row reserve exhausted" <log-file>

Adapt the snippet to your framework. The same pattern holds for PyTorch Lightning, Hugging Face Trainer, DeepSpeed, Megatron-LM, and vLLM training wrappers. Where the wrapper exposes a config flag (for examplelr_scheduler_type in Trainer), prefer the flag over the imperative API to keep the schedule declarative and reproducible.

Select the traceback or open the failed terminal, then run Denpex locally to see the initiating rank, collateral failures, exact fix, and verification command without uploading the log.