KV Cache Corruption on Non-P2P Topologies
KV cache corruption or NaNs in key_value tensors occurring during decoding on hardware topologies that do not fully support PCIe P2P memory sharing.
KV cache corruption or NaNs in key_value tensors occurring during decoding on hardware topologies that do not fully support PCIe P2P memory sharing.
What this failure is
KV Cache Corruption on Non-P2P Topologies is a Hardware failure seen during ML training runs. KV cache corruption or NaNs in key_value tensors occurring during decoding on hardware topologies that do not fully support PCIe P2P memory sharing. Common tags: Vllm, Kv Cache, P2p, Corruption.
Is this what broke your run? Paste your log.
You're reading about KV Cache Corruption on Non-P2P Topologies. Paste your own crash log or traceback below and get the real root cause for YOUR run, not this generic entry. No account, no card. Logs are masked at ingress and never saved to account history.
Want 14 days on the Scale plan?
Request an evaluation code. A verified workplace organization activates up to 50 diagnoses a day, alerts, history, and follow-up questions. No credit card or automatic subscription.
Why it happens (the mechanism)
The software assumes P2P memory access is available and fast, but the underlying topology routes it through the host or blocks it entirely, leading to silent memory corruption. Taken together, these mechanisms explain why the failure is reproducible, why it tends to surface on specific workloads or scales, and why generic mitigation attempts often fall short without addressing the underlying cause.
What you'll observe
- Model outputs NaNs
- KV cache corrupted
Common symptoms and what they mean
| Symptom | Why it happens |
|---|---|
| Incorrect output generated during decoding | The software assumes P2P memory access is available and fast, but the underlying topology routes it through the host or blocks it entirely, leading to silent memory corruption. |
| KV cache mismatch errors | The software assumes P2P memory access is available and fast, but the underlying topology routes it through the host or blocks it entirely, leading to silent memory corruption. |
| nan in key_value tensors | The software assumes P2P memory access is available and fast, but the underlying topology routes it through the host or blocks it entirely, leading to silent memory corruption. |
Which systems are affected
- VLLM
- TensorRT-LLM
- Multi-GPU nodes without NVSwitch or with PCIe ACS enabled
How to confirm this is the problem
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.
- ✓Reproduce the failure from a clean checkpoint/seed: the symptom must appear without warm-up state from a previous run.
- ✓Verified signal present: Incorrect output generated during decoding
- ✓Verified signal present: KV cache mismatch errors
- ✓Verified signal present: nan in key_value tensors
- ✓A targeted fix from the "How to fix it" section eliminates or substantially reduces the symptom within one validation pass.
The fix and the prevention pattern
The root cause is on this page and stays free. A free account adds the exact remediation steps, saved history, and the fix on every entry in the encyclopedia.
Sign up free. Unlock the full analysisNo credit card. Daily allowance follows verified trust tier. Instant access.
Diagnose this failure in VS Code
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.
Install the free VS Code extensionvLLM errors in context
vLLM failures cross request, model, KV cache, worker, kernel and host boundaries. The hub compares common literal signatures and the first control that separates their causal owners.
Compare every vllm error side by sideRelated failures to investigate next
Root cause
- The software assumes P2P memory access is available and fast, but the underlying topology routes it through the host or blocks it entirely, leading to silent memory corruption.
The fix and how to prevent it
Evaluate Denpex on your own logs
Request a Scale evaluation code. A verified workplace organization activates 14 days with up to 50 diagnoses a day. Every account keeps its current diagnosis allowance and gets a verification path. No card or automatic subscription.
Don't just read the fix, diagnose your run
The encyclopedia tells you what went wrong. Denpex tells you what went wrong in YOUR training run. With your logs, your config, and your stack.