ISBC AI-HPC R&D Cloud AMI - Red Hat Enterprise Linux 8.10

This is a repackaged open-source software product with additional charges for ISBC fully managed AI-HPC support. The AMI is designed to function as a head node, pre/post processing node, or compute node within enterprise AI and HPC clusters running on AWS.

• Web-Based HPC Management Console - Monitor and control your entire cluster from a single browser interface, eliminating the need for command-line-only administration and reducing operational overhead for cluster operators.
• Pre/Post Workload Session Management - Manage interactive visualization and data preparation sessions before and after batch jobs, enabling researchers to prepare inputs and analyze outputs without leaving the cluster environment.
• License Monitoring - Track commercial software license usage across your cluster in real time, preventing job failures due to license exhaustion and optimizing license spend.
• SLURM Job Scheduling - Pre-configured SLURM workload manager for multi-node job orchestration, queue management, and resource allocation across GPU and CPU instances.
• Docker Container Orchestration - Run containerized AI/ML frameworks and R&D applications with GPU passthrough support, simplifying software dependency management.
• GPU-Accelerated MLOps with JupyterNotebook - Integrated Jupyter environment with GPU access for interactive model development, training, and inference workflows.
• eVDI (3D) Remote Visualization - Deliver hardware-accelerated 3D visualization sessions to remote users for CAE, CFD, and molecular modeling workloads.
• BYOL Commercial Software Integration - Bring your own licenses for commercial R&D applications and integrate them with the open-source HPC stack without manual configuration.

An engineering team running computational fluid dynamics (CFD) simulations can deploy this AMI as a head node on a general-purpose instance and scale compute nodes on GPU-accelerated instances. Researchers use eVDI sessions for 3D post-processing visualization, while SLURM manages batch job queues across the cluster. License monitoring ensures commercial solver licenses are allocated efficiently across concurrent jobs.

This AMI can be launched on AWS EC2 instances. It is designed to operate within a VPC with appropriate security groups and placement groups configured for low-latency inter-node communication. The head node coordinates job scheduling while compute nodes execute workloads.

• Red Hat Enterprise Linux 8.10 base operating system
• Pre-installed HPC management stack
• SLURM workload manager
• Docker runtime with GPU support
• JupyterNotebook with GPU acceleration
• eVDI session management
• License monitoring dashboard
• ISBC fully managed support

The AMI is optimized for deployment on AWS EC2 instances including GPU-accelerated instance families. It supports integration with Amazon EBS for storage, Amazon VPC for network isolation, and placement groups for HPC-grade inter-node bandwidth.

ISBC provides end-to-end technical support, integration, migration, and maintenance services for enterprise AI and HPC environments. The managed support covers cluster setup, performance optimization, and ongoing operational assistance.

To schedule an HPC architecture consultation or request a pilot deployment, contact ISBC through the support channels listed below.