die-to-die thermal delta between primary die and secondary die emergency throttle

A Blackwell B200 module detected an unsafe temperature difference between its two dies and entered emergency throttling. The delta points to uneven cooling or contact, not ordinary workload-wide thermal saturation. This entry explains how to confirm the cause, apply the fix, and separate it from adjacent blackwell failures.

The literal signature is "die-to-die thermal delta between primary die and secondary die emergency throttle". It is a hardware failure associated with Blackwell, Grace Hopper, NVLink fabrics, and liquid cooling. The line identifies the failing operation or subsystem, while the surrounding evidence decides whether it is the initiating fault or a downstream symptom.

A Blackwell B200 module detected an unsafe temperature difference between its two dies and entered emergency throttling. The delta points to uneven cooling or contact, not ordinary workload-wide thermal saturation. 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 "die-to-die thermal delta between primary die and secondary die emergency throttle".
• 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.

• Blackwell, Grace Hopper, NVLink fabrics, and liquid cooling
• production-shaped multi-accelerator workloads
• containerized and bare-metal deployments of the same stack

Apply the following checklist to a small reproduction: each box below is a positive signal that you are looking at this exact failure rather than a sibling in the same taxonomy.

• ✓Find the first occurrence of "die-to-die thermal delta between primary die and secondary die emergency throttle" and preserve at least 100 lines before it.
• ✓Identify which rank, node, device, or process emitted the first related warning.
• ✓compare both die sensors under idle and a controlled load, then inspect cold-plate contact, coolant flow, and module seating through the approved service path.
• ✓Repeat the same input after the targeted change and require the signature to disappear.
• ✓Resume from latest checkpoint on a thermally healthy GPU only after the control passes.

die-to-die thermal delta between primary die and secondary die emergency throttle

Timestamps and exact values vary across runs, but the pattern. An info-level start, an early WARN, an ERROR carrying the symptom. Is the actual fingerprint you should alert on. The Denpex platform flags this combination automatically.

drain the GPU and preserve per-die temperature, clock, power, and cooling telemetry before changing load or power state. This changes the condition in the causal diagnosis instead of hiding the outer exception. The repeated control proves ownership before recovery from latest checkpoint on a thermally healthy GPU.

# 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 -- "die-to-die thermal delta between primary die and secondary die emergency throttle" <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.

• A Blackwell B200 module detected an unsafe temperature difference between its two dies and entered emergency throttling. The delta points to uneven cooling or contact, not ordinary workload-wide thermal saturation.
• The decisive evidence is the first log line that precedes "die-to-die thermal delta between primary die and secondary die emergency throttle" and differs from a healthy run.
• A nearby failure remains a competing hypothesis until a control separates configuration, capacity, transport, and hardware causes.