Please upload the crash logs. You can’t train on your spark running out of resources and expect things will be ok. You will eventually experience crashes in that situation. Maybe you’re training above your limits. Please share the use case and if possible a repository with the scenario.
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| DGX Spark Shutdown around 95°C during nanoChat Pretraining (20-30 min) | 21 | 2154 | March 23, 2026 | |
| DGX Spark Performance Degradation - GPU Power Draw Issue | 69 | 5355 | June 15, 2026 | |
| DGX Spark stability / out of RAM / overheating | 51 | 3771 | September 7, 2026 | |
| My DGX Spark Hangs ... is this normal? | 5 | 972 | July 15, 2026 | |
| Spark abruptly shuts down | 33 | 1845 | September 3, 2026 | |
| System crashes when memory is full | 35 | 2970 | June 15, 2026 | |
| My DGX Spark keeps freezing and crashing when I try to run this code no matter the LLM | 1 | 583 | December 31, 2025 | |
| Hardware issue | 8 | 751 | December 31, 2025 | |
| DGX Spark becomes unresponsive (“zombie”) instead of throwing CUDA OOM | 16 | 2210 | April 10, 2026 | |
| DGX Spark GB10: partnerdiag PowerStress reproducibly hard-powers-off the box — acpitz 88→97.8°C in 5s, all other field tests pass | 9 | 608 | July 24, 2026 |