Research Gateway Integration with AWS Parallel Computing Service

Research Gateway is an enterprise-grade self-service portal that gives researchers seamless access to scalable high-performance computing (HPC) resources on AWS Cloud. With its integration of AWS Parallel Computing Service (PCS), researchers can easily provision clusters, submit jobs, and track costs – all through a unified interface.

This video demonstrates how Research Gateway integrates with AWS PCS to streamline research workflows, from setting up clusters to visualizing results.

In this video, you will discover:

Use Cases

• Tightly Coupled Workloads: Run MPI applications like CAE, weather modeling, and seismic simulation.
• Accelerated Computing: Use GPUs, FPGAs, and AWS silicon for scientific models and protein prediction
• Interactive Workflows: Prepare, simulate, visualize, and refine results in real time.
• High-Throughput Computing: Power Monte Carlo simulations, image processing, and genomics analysis.

Project Creation & Catalog Integration

• Principal Investigators can quickly create projects in Research Gateway.
• Catalog offers Quick Launch and Advanced PCS templates to meet different research needs.

Cluster Provisioning

• Use AWS PCS Quickstart template for rapid setup.
• Advanced template provides fine-grained control over:
  
  • Cluster size (Slurm scheduler)
  • Number of login & compute nodes
  • Instance types
  • File system options (EFS or FSX)

Job Submission & Monitoring

• Researchers can access clusters securely via SSH.
• Submit jobs with Slurm commands (sbatch, sinfo).
• On-demand compute nodes are spun up automatically to process workloads.

Visualization & Results

• Add visualization nodes from the catalog for GUI-based analysis.
• Pre-configured machine images enable easy installation of visualization tools.

Cost Tracking & Governance

• Research Gateway automatically tracks costs.
• Provides detailed breakdowns by project and cluster for transparency.

Why It Matters

With Research Gateway’s integration of AWS PCS, organizations can:

• Accelerate scientific and engineering research.
• Optimize cloud usage with built-in governance.
• Empower researchers with self-service HPC resources.
• Ensure cost efficiency and compliance.