In addition to providing free access to shared computing resources, the Research Computing Data Core (RCDC) offers researchers the opportunity to obtain prioritized computing capacity and dedicated storage through the Cluster Partnership Program.

The program is designed for faculty who require additional computing or storage capacity or who need prioritized access to cluster resources. Through the program, a Principal Investigator (PI) can purchase professionally managed compute nodes with specifications comparable to those available on the shared cluster. The purchased nodes are integrated into the shared cluster infrastructure, while the PI's research group receives priority access based on the computing capacity contributed by those nodes.

Priority Computing Access

The Cluster Partnership Program provides participating research groups with priority access to cluster resources based on their annual compute contribution. For example, if a PI purchases four CPU-only nodes, with each node containing 64 CPU cores, the annual compute contribution is:

4 nodes × 64 cores/node × 8,760 hours/year = 2,242,560 core-hours/year

This compute contribution forms the basis for the PI's group's annual priority allocation for the lifetime of the contributed nodes, subject to the policies of the Cluster Partnership Program.

In practice, jobs submitted by participating groups receive higher scheduling priority and may start ahead of jobs submitted by users who are not participating in the Cluster Partnership Program or who have exhausted their available priority allocation.

As jobs run, the corresponding priority Service Units (SUs) are consumed. Once a group's priority allocation has been exhausted, subsequent jobs continue to run normally but are scheduled according to the regular cluster scheduling priorities. The priority allocation is automatically renewed annually, subject to program policies.

Priority access can also be customized to support designated users identified by the PI.

Dedicated Storage

Faculty may also purchase dedicated, professionally managed high-performance cluster storage for a specified capacity and term. This provides research groups with additional storage capacity within the professionally managed cluster environment while maintaining the performance, reliability, security, and availability of the RCDC infrastructure.

Professional Infrastructure and Support

RCDC works with top-tier technology vendors to provide an expandable, professionally managed computing environment for the University of Houston research community. The infrastructure encompasses computing hardware, high-performance networking and interconnects, storage, software, systems administration, technical support, and user support.

The Cluster Partnership Program enables research groups to convert research funding into professionally managed computing and storage resources in a highly efficient and cost-effective manner. RCDC supports participating groups throughout the technology lifecycle, including:

• Resource planning: Working with research groups to determine the compute and storage resources that best meet their research needs.
• Competitive procurement: Obtaining competitive pricing from top-tier vendors, including leveraging institutional and bulk-purchasing opportunities.
• Data center integration: Installing and integrating purchased compute resources into RCDC's secure, professionally managed on-campus data center environment.
• Ongoing management: Providing system administration, network administration, software and hardware updates, monitoring, maintenance, and technical support throughout the resource lifecycle.
• Access management: Customizing access and priority policies to support the PI's research group and designated users.

Leverage the RCDC Environment

By integrating purchased resources into the RCDC computing environment, participating research groups can leverage the infrastructure and expertise already available through RCDC, including high-performance network interconnects, shared storage, research software environments, security infrastructure, and professional system administration.

This approach allows research groups to expand their computing and storage capacity without having to independently purchase, deploy, operate, and maintain a dedicated computing environment. Researchers can focus on their science while RCDC provides the underlying infrastructure and professional support required to keep their computing resources secure, reliable, and operational.