We are excited to announce that the Mass Open Cloud, Red Hat, and IBM Research are collaborating to support the new IBM-AI Alliance NAIRR Pilot Deep Partnership projects. With the backing of the US National Science Foundation, the program will provide computing resources and open source AI assets to NAIRR pilot researchers and educators to advance science and education and promote open collaboration in developing and deploying AI.
The rolling submission period began on August 15 and will close October 1, 2025. Projects should be scoped to end by July 1, 2026. (See full submission requirements on the NAIRR pilot website.) Proposals with either research or educational goals can be submitted to one of two tracks:
- Track 1: Core AI projects
- Track 2: AI for science projects
Available resources
Mass Open Cloud: As part of the Deep Partnership program, the MOC provides facilitation support for users and projects, with integration and development support for those new to AI/ML and Kubernetes-style resource management. All operations software is open source, giving experimenters access to the lowest levels of the software stack as needed. Telemetry is stored and available to researchers, and power consumption statistics are available on request for those researching sustainability. Note: The MOC is not certified for use with PII or HIPAA data.
The MOC allocates resources as needed to research projects from a pool of approximately 1,000 servers with 30,000 cores. Ceph storage of up to 50PB is available for project data. Approximately 270 Nvidia GPUs (mostly A100SXM4s and H100s) and a small number of AMD MI210 accelerator graphics cards are available. The MOC provides high-speed networking to public and research infrastructure, along with available external IP addresses as needed. Details on usage pricing for various resources are available in New England Research Cloud documentation. The resources provided to the NAIRR pilots will be allocated as usage credits, with a maximum quota specified for the credits. Researchers may choose how to allocate their credits among resources.
Red Hat AI open cloud platform: The Red Hat AI software stack includes Red Hat Enterprise Linux (RHEL) and OpenShift AI for enterprise application development. This environment provides tools across the full lifecycle of AI/ML experiments and models and helps build, train, test, and deploy models optimized for hybrid cloud environments.
Open Source AI Models & Tools from IBM Research: Available resources differ according to track (find more details in the AI Alliance-NAIRR Pilot Overview).
- Core AI Track:
- IBM Granite Family of Models (Language, Code, Embedding, Speech, Vision, Guardian, Geospatial, Time series)
- Instruct Lab (for improving LLM alignment)
- Docling (AI-driven document conversion)
- Data Prep Kit (for code and language modalities)
- Unitxt (Python library for LLM evaluation)
- Risk Atlas Nexus (for AI governance workflows)
- EvalAssist (LLM-as-a-Judge framework)
- AI Attribution toolkit
- AI for Science Track:
- Foundation Model for Materials (FM4M)
- Biomedical Foundation Models (BMFM)
- IBM-NASA and IBM-ESA Foundation Models for Earth (Prithvi and TerraMind)
The Red Hat Research team is excited to be part of this effort to bring the benefits of open source to the development of AI research and education. Open source development is a proven model for accelerating innovation, expanding access, improving security, and building a vibrant AI ecosystem. We invite you to apply to this program at the NAIRR Deep Partnership site, and we look forward to sharing the chosen projects and their results in the months to come.
*Header image credit: US National Science Foundation
