AWS adds SOCI indexing to speed container startup for SageMaker AI training jobs
Overview
AWS announces Seekable OCI (SOCI) indexing for Amazon SageMaker AI training and processing jobs, which lets containers start without waiting for the full image download.
In AWS benchmarks on eight images from 1.8 GB to 35.4 GB, lazy loading took 3 to 11 seconds, compared with 48 seconds to 7 minutes 17 seconds for standard downloads. AWS says training performance is unaffected beyond a first-epoch penalty.
The feature works only with Amazon ECR private repositories and is available in all AWS Regions. Jobs fall back to a standard download when needed, and AWS says a SOCI index never prevents a job from running.
Written by AI from the articles below · updated Oct 27, 1:35 PM ET
Check the sources:
Article timeline
The articles in this story. Times are ET.
- AWS Machine Learning BlogOfficialSageMaker AI adds SOCI indexing to cut container image download time for training jobs
AIAWS announces Seekable OCI (SOCI) indexing for Amazon SageMaker AI training and processing jobs, which lazy-loads container images and starts containers without waiting for the full download. In AWS benchmarks on eight images from 1.8 GB to 35.4 GB, lazy loading took 3–11 seconds, against 48 seconds to 7 minutes 17 seconds for standard downloads. AWS says training performance is unaffected beyond a first-epoch penalty.
Heat Trend
- Current comparable heat
- 6
- Peak comparable heat
- 10Oct 27, 2:00 PM ET
- Change in past 24 hours
- –
The trend compares only continuously observed participants, so its scope may be smaller than the current heat total. Hover or tap for hourly heat; use arrow keys with a keyboard.