The Athena GPU Cluster is a high-performance computing (HPC) resource available to Technion researchers and students.
It is designed to accelerate computational workloads, including artificial intelligence (AI), machine learning (ML), and scientific computing.
Cluster Overview
- The cluster consists of nine Nvidia DGX compute nodes, each equipped with eight Nvidia A100 GPUs.
- All nodes run a GNU/Linux operating system.
- Access to Athena is provided through the login server:
dgx-master.technion.ac.il. - Jobs are executed using containerized workflows and managed via the SLURM scheduler.
Using Athena
The Athena Cluster requires users to work with containers to ensure software portability and reproducibility.
Users typically download containers from Nvidia NGC
and submit their workloads through SLURM.
For a step-by-step guide on accessing Athena, submitting jobs, and managing containers, see the
User Instructions.
Getting Started
- HPC Account Registration: A valid HPC account is required to access Athena. Register here.
- Billing & Pricing: Athena is a paid resource, and GPU usage is billed per hour. See the Billing Procedure.
- Support: For technical issues, visit the Support page.
- Hardware Specifications: Learn more about the cluster’s architecture here.
Additional Resources
- Training Sessions: Participate in Nvidia AI experts monthly sessions to enhance your skills.