Announcing instance preference lists for Amazon SageMaker AI training jobs
Amazon SageMaker ยท 2026-09-15
Actions
Technical Details
| Regions | all |
|---|---|
| Cost Impact | Neutral |
What This Means
For DevOps Teams
Update your Amazon SageMaker Python SDK and configure instance preferences in your Compute settings to automatically secure available GPU capacity, reducing manual retries and operational overhead.
For Platform Teams
Adopt instance preference lists to simplify GPU capacity management, enhance resource utilization, and reduce the need for custom retry scripts, leading to more efficient AI training workflows.
For Executives
Implement instance preference lists to reduce wait times for GPU resources, enabling faster model experimentation and development, and improving overall team productivity.
Source
Related Amazon SageMaker Updates
- Qwen3.6-35B-A3B-NVFP4 and Wan2.1-T2V-1.3B-Diffusers models now available on Amazon SageMaker JumpStart (2026-09-14)
- Ministral-3-3B-Instruct-2512 and Ministral-3-8B-Instruct-2512 models now available on Amazon SageMaker JumpStart (2026-09-14)
- Amazon SageMaker HyperPod now supports model caching for faster inference autoscaling and reduced cold starts (2026-09-11)
- CVE-2026-83551 - Cleartext storage of HMAC signing key in Amazon SageMaker Python SDK (2026-09-09)
- Amazon SageMaker Feature Store introduces UpdateRecord for feature-level writes (2026-09-08)