Uploaded May 2026 | Updated September 2026, 2 weeks ago

---

Checkpointing is a critical component of large-scale AI training, enabling resilience, efficiency, and flexibility in modern machine learning workflows. We will explore the issues that come with the scaling up of training model sizes and take a deep dive into the workload pattern of checkpointing and how the right storage can reduce the downtime of AI training.

Presented by John Mazzie | Micron Technology - Member of Technical Staff, Systems Performance Engineer

Learn More:
• SDC: StorageAI Website: snia.org/sniadeveloper/storageai
• SNIA Website: snia.org
• SNIA Educational Library: snia.org/library
• X: twitter.com/SNIA
• LinkedIn: linkedin.com/company/snia