{"version":1,"type":"story","url":"https://digestai.news/story/amazon-sagemaker-ai-adds-instance-preference-lists","json":"https://digestai.news/story/amazon-sagemaker-ai-adds-instance-preference-lists.json","markdown":"https://digestai.news/story/amazon-sagemaker-ai-adds-instance-preference-lists.md","slug":"amazon-sagemaker-ai-adds-instance-preference-lists","headline":"Amazon SageMaker AI Adds Instance Preference Lists","summary":"AWS has introduced **Instance preference lists** for Amazon SageMaker AI Training Jobs and Processing Jobs, allowing users to specify a list of up to five instance types in order of preference. When a job is submitted, the system automatically checks available capacity from the user's preferred list and launches on the first type with availability. This feature reduces wait times and manual retry loops, enabling faster job starts and more time for model development. Customer challenges often arise during peak demand periods when preferred GPU instances are not immediately available. The solution provided by this feature is to let Amazon SageMaker AI automatically evaluate multiple instance types at job creation, ensuring that jobs start promptly on the first available type or fall back to alternative instance types using on-demand capacity if needed. This removes the need for manual retry scripts and adds operational overhead.\n\n**Key Points:*\n- Users can specify up to five preferred instance types in order of preference.\n- Amazon SageMaker AI automatically evaluates the list in priority order and launches the job on the first available type.\n- The feature reduces wait times and allows teams to focus more on model development rather than managing constraints.\n\n**Why It Matters:**\nThe introduction of Instance preference lists for Amazon SageMaker AI Training Jobs is significant as it addresses a common challenge faced by users who need access to specific GPU configurations during peak demand periods. This simplifies the process, reduces operational overhead, and speeds up job starts, making it easier for teams to iterate on their models.\n\n**Category:**\nModels\n\n**Entities:**\n- **Companies:** AWS, Amazon SageMaker\n- **Models:** Not mentioned in the article\n- **People:** Not mentioned in the article\n\n**Content Type:**\nNews\n\n**Importance:**\n2\n\n**Is AI News?**\ntrue\n\n**Model Release:**\nnull\n\n**Funding:**\nnull","keyPoints":["Users can specify up to five preferred instance types in order of preference.","Amazon SageMaker AI automatically evaluates the list in priority order and launches the job on the first available type.","The feature reduces wait times and allows teams to focus more on model development."],"whyItMatters":"This feature simplifies the process for users who need specific GPU configurations during peak demand periods, reducing operational overhead and speeding up job starts.","category":{"slug":"models","name":"Generative AI & Models","url":"https://digestai.news/category/models"},"entities":{"companies":["AWS","Amazon SageMaker"],"models":[],"people":[]},"firstPublishedAt":"2026-09-15T16:01:47Z","updatedAt":"2026-09-15T16:01:47Z","sourceCount":1,"hasPrimarySource":true,"sources":[{"outlet":"AWS Machine Learning Blog","title":"Announcing instance preference lists for Amazon SageMaker AI training jobs","url":"https://aws.amazon.com/blogs/machine-learning/announcing-instance-preference-lists-for-amazon-sagemaker-ai-training-jobs","publishedAt":"2026-09-15T16:01:47Z","type":"primary","primary":true,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Amazon SageMaker AI Adds Instance Preference Lists\", 15 September 2026, https://digestai.news/story/amazon-sagemaker-ai-adds-instance-preference-lists","publisher":"Digest AI","title":"Amazon SageMaker AI Adds Instance Preference Lists","datePublished":"2026-09-15T16:01:47Z","url":"https://digestai.news/story/amazon-sagemaker-ai-adds-instance-preference-lists"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}