AI Supercomputer Access Request

Description

The AI Supercomputer Access Request allows UAlbany researchers to request access to the on-premises NVIDIA DGX cluster for GPU-intensive machine learning, deep learning, and related research computing workloads. The environment supports use of Jupyter notebooks, JupyterLab, Python scripts, and supported NVIDIA container-based workflows.

What’s Included

  • Access to 24 NVIDIA DGX A100 systems, each with 8 NVIDIA A100 GPUs
  • Support for GPU-intensive artificial intelligence, machine learning, deep learning, model training, inference, and related research computing workloads
  • Support for Jupyter notebooks, JupyterLab, Python scripts, and supported NVIDIA container-based workflows
  • Up to 1 TB of high-performance, all-flash storage for data actively being used by AI Supercomputer workloads
  • A standard 10 TB allocation of disk-based research storage for the research group’s general data storage needs
  • Non-priority GPU access subject to the job, GPU, runtime, scheduling, and preemption limits documented in the AI Supercomputer Service Level Agreement
  • Access to self-service documentation, tutorials, and instructions for connecting to the system and submitting workloads

What’s Not Included

Eligibility

UAlbany faculty and researchers may request access for university research. A supervising faculty member must request or sponsor access for graduate students, postdoctoral researchers, and other members of their research group.

More Information