AWS re:Invent Recap: Habana Gaudi Based EC2 System

What Happened: 

With machine learning workloads growing rapidly and faster machines needed for training models in the cloud, Andy Jassy announced Habana Gaudi-based Amazon EC2 Instances will be available in the first half of 2021. Powered by new Habana Gaudi pre-processors from Intel, users can expect a 40% better price/performance over the current GPU based EC2 instances. Built specifically for ML training, these instances work seamlessly with TensorFlow and Pytorch for training deep learning models.

Why It’s Important: 

  • Competitive Edge: This has solidified AWS as the top choice for cloud-based systems suitable for large Machine Learning workloads. 
  • Cost: With 40% better price/performance, organizations, and partners will benefit from significant cost savings.
  • Easily Integrated: using Habana Gaudi instances integrates with and natively supports common frameworks already used such as TensorFlow and PyTorch.
  • Seamless Transition: These instances are used with familiar tools and technology like AWS Deep Learning AMIs, Amazon EKS and ECS for containerized applications, and Amazon SageMaker.

Why We’re Excited

Developers now have the power to build new training models or port existing models from graphics processing units to Gaudi accelerators using Habana’s SynapseAI Software Suite. This reduces the cost of training AI models at scale, allowing for a rapid and sophisticated training process that results in a faster, more optimized delivery of solutions.  


Habana Gaudi EC2 instances will be available in the first half of 2021. Sign up here for early access.

If you’re looking to explore these services further and need some guidance, let us know and we’ll connect you to an Idexcel expert!