←View full SambaNova status
SambaNova: Inference Timeout / Model Loading Error
Current Status: Major Outage
Last checked: 7m ago
What We're Seeing Right Now
No recent issues reported. If you're experiencing problems with SambaNova, report below to help the community.
What is this error?
When SambaNova inference times out, the model took too long to load, initialize, or generate a response. Large models can have cold start times of 30-120 seconds, and inference itself can timeout under load.
Error Signatures
Inference timeoutModel loadingCold start504 Gateway TimeoutRequest timed outModel initialization failedPrediction timed outWorker not readyCommon Causes
- Cold start — model loading into GPU memory
- Model is too large for allocated resources
- Input is too large or complex
- Infrastructure overloaded
- SambaNova inference endpoint is degraded
✓ How to Fix It
- Increase timeout values in your client
- Use a smaller model variant if available
- Keep the endpoint warm with periodic requests
- Check if auto-scaling is configured
- Reduce input size
- Check this page for infrastructure issues
Live Signals
Service Components
SambaNova API
OperationalRecent Incidents
No incidents in the past 30 days
Frequently Asked Questions
Why is SambaNova inference timing out?
Large models have cold starts (30-120s). If timeouts persist, the model may need more resources or SambaNova may be overloaded.
How do I reduce SambaNova cold start time?
Keep endpoints warm, use smaller models, or use SambaNova's dedicated/reserved infrastructure.
Is SambaNova inference slow for everyone?
Check community reports below for real-time performance feedback.