Relevance Lab GPU-Optimized Windows Data Science AMI for AWS RES

This Windows Server 2022 Amazon Machine Image (AMI) by Relevance Lab combines CIS Level 1 security hardening, full AWS Research and Engineering Studio (RES) compatibility, and GPU optimization in a single, ready-to-deploy image. Instead of spending days manually installing and configuring data science tools, hardening the OS, and validating compliance, your team can launch a fully functional workspace and begin productive work within minutes of instance launch.

Most data science AMIs offer either a tool stack or security hardening - not both. This workspace uniquely delivers:

• CIS Level 1 compliance validated via Amazon Inspector - reducing your team's audit preparation burden and satisfying security requirements without manual hardening
• Native AWS RES integration - enabling centralized session management, cost tracking, and multi-user deployment that standalone AMIs cannot provide
• Full GPU acceleration stack - NVIDIA drivers, CUDA Toolkit, and cuDNN pre-installed and tested, eliminating driver compatibility troubleshooting

This AMI uses a metered billing model with software charges applied on top of standard EC2 infrastructure costs. Visit the Pricing tab on this listing for exact rates by instance type. No free trial is currently available, but a structured pilot program is offered - contact the Relevance Lab team to arrange a guided evaluation.

• Security and Compliance: OS hardened to CIS Level 1 Benchmarks and validated with Amazon Inspector, reducing your compliance verification effort and satisfying audit requirements for regulated research environments
• RES Compatibility: Seamlessly integrates with AWS Research and Engineering Studio for centralized session management, user provisioning, and cost allocation across teams
• Remote Desktop Access: High-performance GUI powered by NICE DCV, delivering responsive remote desktop experience for graphical workloads
• GPU-Ready: NVIDIA drivers, CUDA Toolkit, and cuDNN pre-installed and configured for immediate deep learning and HPC workloads on GPU instance families (G4dn, G5, P4d)
• Development Environments: VS Code, Visual Studio 2022 CE, PyCharm CE, and RStudio - ready to use without installation or configuration
• Data Science and ML Tools: JupyterLab with Python and R kernels, PyTorch, TensorFlow, scikit-learn, PySpark, Dask, and Vowpal Wabbit
• Container and Environment Management: Docker, Docker Compose, and Anaconda for reproducible workflows
• Productivity Suite: LibreOffice for documents, spreadsheets, and presentations
• Web and Utility Tools: Chrome, Git, 7-Zip, AWS CLI

For GPU-accelerated workloads: G4dn, G5, or P4d instance families. For CPU-only workloads: M5, C5, or R5 families. Minimum recommended: 16 GB RAM, 4 vCPUs, 100 GB EBS storage.

Contact the Relevance Lab team to schedule a guided deployment walkthrough or request a structured pilot for your research group. The pilot allows your team to validate the full workflow - from AMI launch through RES integration to running GPU-accelerated training jobs - before committing to a broader rollout.

• Operating System: Windows Server 2022 (CIS Level 1 Compliant)
• Remote Access: Amazon NICE DCV (port 8443)
• Languages: Python 3.x, R
• IDEs: VS Code, Visual Studio 2022 CE, PyCharm CE, RStudio
• Notebooks: Jupyter Notebook, JupyterLab
• Frameworks: PyTorch, TensorFlow, scikit-learn, PySpark, Dask, Vowpal Wabbit
• Office Tools: LibreOffice (Writer, Calc, Impress)

1. Subscribe to this AMI from the AWS Marketplace listing
2. Launch on a supported EC2 instance type (G4dn, G5, or P4d for GPU workloads)
3. Configure your security group to allow inbound TCP on port 8443 for NICE DCV access
4. Connect via NICE DCV at https://[instance-public-ip]:8443
5. Verify GPU availability by opening a terminal and running nvidia-smi
6. For RES deployments, register the AMI in your RES environment for managed session provisioning

Expected time-to-value: your team can be running notebooks and training models within minutes of instance launch.

A deployment guide covering RES registration, GPU verification steps, and a sample PyTorch notebook demonstrating the pre-configured ML stack are available. Visit the Additional Resources section on this listing page for direct access to documentation.

AI/ML professionals, research teams, data analysts, and developers in regulated or security-conscious environments who need a GPU-enabled, compliance-ready Windows workspace without the overhead of manual setup and hardening.