Amazon EC2 instances powered by Habana Gaudi
Amazon EC2 instances powered by Gaudi accelerators from Habana Labs, an Intel company, are specifically designed for training deep learning models. The new EC2 instances will leverage up to 8 Gaudi accelerators and deliver up to 40% better price performance than current GPU-based EC2 instances for training deep learning models. Gaudi-based EC2 instances are ideal for deep learning training workloads of applications such as natural language processing, object detection and classification, recommendation and personalization.
The increase in complexity of machine learning models is driving up both the time and cost to train. Customers continue to look for ways to reduce the cost of training in order to iterate more frequently and improve the quality of their ML models. Gaudi-based EC2 instances are designed to address these customer needs by delivering cost efficiency in training machine learning models.
Software for Gaudi accelerators includes a SDK and other tools that simplify building and training models. The SDK will be natively integrated with common ML frameworks like TensorFlow and PyTorch, and provide the ability to easily port existing training models from using GPUs to Gaudi accelerators. Customers will be able to launch these new EC2 instances using AWS Deep Learning AMIs, Amazon EKS and ECS for containerized applications, and Amazon SageMaker.
Sign-up for early access
Habana Gaudi-based EC2 instances will be available in 2021. To be notified about early access, sign up here, and we’ll inform you when more information becomes available.