AWS Batch Adds Quota Sharing for AI Training Jobs
Amazon Web Services has added quota management to AWS Batch for SageMaker Training jobs. Teams or projects can receive set shares of GPU capacity, lend unused capacity to each other automatically, and let top-priority jobs run first, all with less manual setup.
Quick glossaryAmazon Web Services, GPU, preemption, SageMaker
- Amazon Web Services
- The cloud computing arm of Amazon, best known for renting out servers, storage and computing power to businesses.
- GPU
- A computer chip that performs many calculations at once, and is widely used to train and run AI systems.
- preemption
- Pausing a running task so that a more urgent task can use the computing power first.
- SageMaker
- Amazon's platform for building, training and deploying machine learning models.
AWS Machine Learning Blog
Burning Potato score: 65%