Sensitive Data Research Environment (SDRE)

The Sensitive Data Research Environment (SDRE) is the University of Alberta’s secure high-performance digital infrastructure. As a cornerstone of the university’s AI Strategy, the SDRE provides a scalable platform for research involving sensitive data governed by legislation, regulatory frameworks and contractual obligations.

By centralizing security assessments and governance, the SDRE reduces operational overhead for researchers, allowing you to focus on innovation while meeting the stringent security expectations of national funding agencies and security guidelines.

Why choose the SDRE? 

• Security & compliance: Alignment with UAPPOL policies and Government of Canada guidelines for sensitive technology research.
• AI-optimized power: Access to state-of-the-art NVIDIA H200 and L40S GPU nodes.
• Cost efficiency: Eliminates the need for individual research groups to maintain separate, decentralized hardware.
• Seamless collaboration: Securely share data with internal and external collaborators under a unified governance framework.

Systems specifications: built for the future of AI

The SDRE is a high-performance computing (HPC) cluster specifically designed for AI, deep learning and data-driven discovery.

Resource Type	Best For	Hardware Highlights
AI & Deep Learning	Large-scale model training, LLMs, massive simulations	NVIDIA H200 GPUs (56 total), 728 CPU cores, 14TB RAM
Inference & Viz	Model deployment, scientific imaging, real-time inference	NVIDIA L40S GPUs (44 total), 704 CPU cores, 7TB RAM
Data-Driven Research	Genomics, statistics, parallel CPU computing	896 CPU cores, 10.75TB RAM
Storage & Network	Active project data and low-latency communication	Ultra-fast SSD storage & high-speed Infiniband networking

WHO CAN ACCESS THE SDRE?

Access is available to the University of Alberta research community:

• Faculty & Principal Investigators (PIs): Lead research projects requiring high-performance resources or sensitive data handling.
• Graduate students & staff: Work under an approved SDRE project sponsored by a PI.
• External collaborators: Access via sponsored accounts tied to a U of A project.

The SDRE environment is also available to researchers with non-sensitive data, provided they adhere to the environment’s security protocols.

Getting Started: Account & Data Management

1. Requesting an account

To launch a project on the SDRE, Principal Investigators (PIs) must submit a New Project Account Request.

• Requirements: An active RES number, a brief project description, and a list of team CCIDs. NOTE: If you do not have an active RES number, a temporary placeholder number will be generated as part of the New Project Account Request process. Please note that this is for administrative tracking only and must not be confused with the official RES number issued upon the awarding of a grant.
• Consultation: Because the SDRE is a secure environment, IST staff will reach out for an initial consultation to ensure suitability and compliance.

2. Data Storage Hierarchy

The SDRE utilizes a tiered storage approach to balance performance and long-term retention:

• Home (25 GB): For small scripts and configuration files. (Not backed up).
• Project (5 TB): High-performance space for active computational research.
• Research Data Storage Service (RDSS) (50 TB): Long-term storage of data that is not actively being used for computing tasks.

3. Access & Security

Access is managed via a dedicated SDRE VPN  and requires Multi-Factor Authentication (MFA). Users can connect via:

• Secure SSH
• Virtual Client Services (VCS)
• Open On-Demand (Web-based portal — Coming Soon)

4. Data transfer and management

Globus is integrated with the RDSS, allowing for secure data transfer into this storage system. The RDSS serves as a staging area for moving data into the SDRE. Instructions for connecting with the Globus web application and transferring data into the RDSS is available in the U of A User Guide for Transferring Data Using Globus.

For more information, connect with the Digital Research Services team at research.support@ualberta.ca.

Support + training

Comprehensive technical documentation, including Slurm job scheduling, software module management (CVMFS), and Apptainer (container) usage, is available via our Technical Wiki webpage, access to which will be provided during onboarding. Please note that off-campus access to the Wiki requires use of the U of A VPN. 

Digital Research Services offers a variety of workshops aimed at developing skills required to perform intensive computing in a high-performance computing (HPC) environment. Visit the Research Computing Training web page to see upcoming workshops and bootcamps. 

Questions? Contact the Digital Research Services team at research.support@ualberta.ca.