{"id":1125291,"date":"2025-02-17T12:00:59","date_gmt":"2025-02-17T20:00:59","guid":{"rendered":"https:\/\/www.microsoft.com\/en-us\/research\/?post_type=msr-project&p=1125291"},"modified":"2026-03-04T12:37:41","modified_gmt":"2026-03-04T20:37:41","slug":"national-ai-research-resource-nairr-pilot","status":"publish","type":"msr-project","link":"https:\/\/www.microsoft.com\/en-us\/research\/project\/national-ai-research-resource-nairr-pilot\/","title":{"rendered":"National AI Research Resource (NAIRR) Pilot"},"content":{"rendered":"
\n\t
\n\t\t
\n\t\t\t\"\"\t\t<\/div>\n\t\t\n\t\t
\n\t\t\t\n\t\t\t
\n\t\t\t\t\n\t\t\t\t
\n\t\t\t\t\t\n\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\n\n

Microsoft Support for the National AI Research Resource Pilot<\/h1>\n\n\n\n

<\/p>\n\n\t\t\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t<\/div>\n\t\t<\/div>\n\t<\/div>\n<\/section>\n\n\n\n\n\n

Microsoft’s commitment to the NAIRR pilot<\/h2>\n\n\n\n

Microsoft is delighted to contribute as a lead supporter of the National AI Research Resource (NAIRR) pilot. (opens in new tab)<\/span><\/a> The mission of the NAIRR pilot aligns with our commitment to broaden AI research and spur innovation by providing greater computing resources to AI researchers and engineers in academia and non-profit sectors. As part of our commitment to the NAIRR pilot (opens in new tab)<\/span><\/a>, Microsoft has offered $20 million in Microsoft Azure compute credits. Read below for more information on what we’re offering researchers, and check out the tabs above for exciting research directions, eligibility details, and awards.

<\/p>\n\n\n\n

What’s available for researchers<\/h2>\n\n\n\n


<\/p>\n\n\n\n

High-performance computing resources<\/h3>\n\n\n\n

We provide high-performance computing resources to meet various workload requirements. <\/p>\n\n\n\n

For larger workloads<\/em>: NVIDIA A100 GPUs with InfiniBand (opens in new tab)<\/span><\/a> connectivity are available for select NAIRR Pilot projects to develop world-class AI for science. Researchers can request access to NDm_A100_v4 series (opens in new tab)<\/span><\/a> virtual machines (VM) designed for high-end Deep Learning training and tightly coupled scale-up and scale-out HPC workloads. Preference will be given to projects ready to onboard immediately that can run the GPUs at full utilization.<\/p>\n\n\n\n

For smaller workloads<\/em>: Researchers can request NCads_H100_v5 (opens in new tab)<\/span><\/a> SKU series powered by NVIDIA H100 NVL GPU and 4th-generation AMD EPYC\u2122 Genoa processors; NC_A100_v4 (opens in new tab)<\/span><\/a> SKU series powered by NVIDIA A100 PCIe GPU and third generation AMD EPYC\u2122 7V13 (Milan) processors; and Azure HBv2 (opens in new tab)<\/span><\/a> and HBv3 (opens in new tab)<\/span><\/a> series VMs. We recommend the NC series GPUs for AI training workloads where the model can be trained on either a single NC_A100_v4 instance with a maximum of 4 80G A100 GPUs (320 GB GPU memory), or a single NCads_H100_v5 instance with a maximum of 2 94G H100 GPUs (188 GB GPU memory). Proposals can request multiple NC_v4 or NC_v5 instances to run parallel workloads.
<\/p>\n\n\n\n

<\/div>\n\n\n\n

Leading-edge models<\/h3>\n\n\n\n

We’re dedicated to delivering frontier and open models to further research, including popular large language and vision foundation models. Researchers can access these models through our Azure OpenAI Service (opens in new tab)<\/span><\/a> and the Azure AI model catalogue (opens in new tab)<\/span><\/a>. The Azure AI model catalog features over a thousand models (opens in new tab)<\/span><\/a> including models curated by Microsoft, OpenAI, Hugging Face, and more. (Access is available for all catalog models, except those listed as requiring Azure Marketplace.)
<\/p>\n\n\n\n

<\/div>\n\n\n\n

Domain-specific resources<\/h3>\n\n\n\n

We recognize the benefit of domain-specific resources for some research projects. For this reason, we offer health and life science models (opens in new tab)<\/span><\/a>, innovative tools for chemistry and materials science research via Azure Quantum Elements (opens in new tab)<\/span><\/a>, and tools for research and development on AI fairness, accuracy, reliability, and interpretability (opens in new tab)<\/span><\/a>. Researchers can also access datasets including <\/a>Fields of the World (opens in new tab)<\/span><\/a>.
<\/p>\n\n\n\n

<\/div>\n\n\n\n

Scientific support<\/h3>\n\n\n\n

We’re dedicated to helping researchers make the most of these resources. To support domain experts who are new to machine learning, researchers can apply for a research collaboration with Microsoft\u2019s scientists and engineers, including researchers at Microsoft Research and Microsoft\u2019s AI for Good Lab<\/a>.<\/p>\n\n\n\n

<\/div>\n\n\n\n
<\/div>\n\n\n\n\n\n

Are you eligible?<\/h3>\n\n\n\n

To be eligible for awards under the NAIRR pilot, applicants must be US-based researchers and educators from US-based institutions including academic institutions, non-profits, federal agencies or federally funded R&D centers, state, local, or tribal agencies, or startups and small businesses with federal grants. We encourage submissions led by teams located in Established Program to Stimulate Competitive Research (EPSCoR) jurisdictions or rural areas, and from Minority Serving Institutions (MSI).\u00a0<\/p>\n\n\n\n

<\/h3>\n\n\n\n
<\/div>\n\n\n\n\n\n

<\/p>\n\n\n\n


Opportunities to use Azure resources through the NAIRR pilot are still available, by submitting a request for a CloudBank Research award. For more details, see NAIRR Pilot – NAIRR Pilot Resource Requests to Advance AI Research (opens in new tab)<\/span><\/a> and select “CloudBank Research” from the list of Resources.
<\/p>\n\n\n\n

<\/h3>\n\n\n\n
<\/div>\n\n\n\n\n\n

Grand Challenges<\/h3>\n\n\n\n

We’re excited by breakthroughs that can be enabled by leveraging state of the art models for health<\/a>, molecular<\/a> and materials discovery<\/a>. Building on the momentum of breakthroughs to date using these models, Microsoft is eager to share the tools we have for advancing world-class science to support grand challenge research projects through the NAIRR Pilot\u2019s new Deep Partnership track.<\/p>\n\n\n\n

<\/div>\n\n\n\n

<\/p>\n\n\n\n

Grand Challenge Awards<\/h3>\n\n\n\n

We are pleased to recognize the following researchers who were selected as finalists for Azure grand challenge awards. See below for more information on their proposed projects.<\/p>\n\n\n\n

\n\n\n\n

Accelerating Multifunctional C-S-H Seeding Materials Discovery through Agentic AI and Scalable Cloud Computing<\/strong><\/p>\n\n\n\n

Prasanna Venkataraman Balachandran (University of Virginia)<\/em><\/strong><\/strong><\/p>\n\n\n\n

Concrete is the most widely used man-made material globally, with its composition and production methods largely unchanged for over 200 years. Cement production is responsible for nearly 8% of total CO2 emissions. In developed nations, infrastructure repair is a major challenge, and the U.S. alone requires $2.8 trillion for its aging transportation systems. Low-carbon cements could reduce cement emissions by 20-60%, but they struggle with early-age strength development, hindering their adoption. Research on calcium-silicate-hydrate (C-S-H) seeding has shown promise in accelerating cement hydration, highlighting the need for new multifunctional C-S-H seeding materials. This proposal aims to develop an innovative AI pipeline that integrates advanced AI techniques with domain-specific constraints and physics-based simulations to discover and optimize novel C-S-H seeding materials.<\/p>\n\n\n\n

\n\n\n\n

Accelerating molecular design and crystal structure prediction for carbon capture<\/strong><\/p>\n\n\n\n

Jeffrey Neaton (University of California, Berkeley)<\/em><\/strong><\/strong><\/p>\n\n\n\n

Crystal structure prediction (CSP) is a longstanding challenge in materials science, with conventional methods relying on computationally expensive density functional theory (DFT) calculations that limit their applicability to small systems. Recent advances in machine learning interatomic potentials (MLIPs) and generative AI offer a paradigm shift, enabling efficient exploration of vast configuration spaces with near-DFT accuracy at greatly reduced cost. This project aims to develop an integrated AI-driven CSP workflow that combines MLIPs, generative AI, DFT, and classical simulations to predict the crystal structures of complex organic systems, demonstrating our workflow on polyamine molecular crystals as promising candidates for carbon capture. The resulting workflow and transferable MLIPs will be made publicly available, providing powerful new tools for CSP of molecular crystals and accelerating the discovery of materials for energy and environmental applications. <\/p>\n\n\n\n

\n\n\n\n

RxFM: A Multimodal Foundation Model for ALS Therapeutic Science and Drug Discovery<\/strong><\/p>\n\n\n\n

Mark Albers (Massachusetts General Hospital,<\/em><\/strong> Harvard University)<\/em><\/strong><\/p>\n\n\n\n

Amyotrophic Lateral Sclerosis (ALS) is a fatal neurodegenerative disease with no cure and a complex, multifactorial etiology. This project aims to develop a Multimodal Therapeutic Foundation Model (RxFM) that will be trained on diverse biomedical and pharmacodynamic data to accelerate Amyotrophic Lateral Sclerosis (ALS) drug discovery and drug repurposing. This RxFM model will integrate multi-omic, single-cell, metabolomic, clinical, and drug perturbation data within a unified framework. The project aims to extend advances in biologic foundation models to model therapeutic actions, seeking to identify interventions that reverse disease signatures. Moreover, the RxFM framework can be fine-tuned or queried for diseases with little known biology, supporting zero-shot drug discovery.   <\/p>\n\n\n\n

\n\n\n\n

Accelerating Neurodegeneration Discovery with Agentic AI Systems in C-BRAIN<\/strong><\/p>\n\n\n\n

Randall Bateman (Washington University in St. Louis)<\/em><\/strong><\/p>\n\n\n\n

Alzheimer\u2019s disease research faces a critical challenge: despite decades of effort, more than 99% of drug candidates fail in clinical trials. This failure reflects a deeper problem in biomedical science – vital knowledge is scattered across millions of papers, complex datasets, and unpublished \u201cdark data,\u201d making it nearly impossible for a scientist to understand the complete picture.<\/p>\n\n\n\n

The Consortium for Biomedical Research and AI in Neurodegeneration (C-BRAIN) is tackling this challenge by combining the expertise of leading scientists with cutting-edge AI. This project aims to develop and test three unique AI tools that mirror the way science works: generating hypotheses, testing them against data, and providing rigorous peer-review style critique. At each step, human experts will remain in the loop to guide, validate, and refine outputs, ensuring trust and scientific rigor. Together, these AI tools will form the first open benchmark for \u201cAI Scientist\u201d systems; thereby, allowing the broader research community to compare and improve their own approaches. The goal is to deliver a transparent, end-to-end pipeline that will accelerate discovery in Alzheimer\u2019s and set the stage for scalable, AI-enabled science across biomedicine.  <\/p>\n\n\n\n

\n\n\n\n

Braingents: Developing a Longitudinal Agentic AI Framework for Alzheimer’s Disease Neuroimaging Biomarkers<\/strong><\/p>\n\n\n\n

Ish Talati (Stanford University School of Medicine)<\/em><\/strong><\/strong><\/p>\n\n\n\n

This project aims to develop an agentic AI framework that autonomously orchestrates end-to-end neuroimaging analysis while preserving expert-level accuracy. The framework will bring together four specialized agents: a Data Harmonization Agent that normalizes scanner and protocol variability across sites using domain adaptation; a Multi-Modal Detection Agent leveraging vision transformers to quantify structural atrophy and amyloid\/tau burden; a Longitudinal Tracking Agent for precise monitoring of hippocampal volumes, cortical thinning, and plaque progression; and a Clinical Reasoning Agent that generates interpretable, language model\u2013driven reports. By combining automation, interpretability, and multimodal scalability, the research aims to accelerate biomarker discovery while paving the way for clinical readiness in Alzheimer\u2019s disease neuroimaging.<\/p>\n\n\n\n

\n\n\n\n

Trustworthy and Secure Cloud-Native Agentic OpenROAD<\/strong><\/p>\n\n\n\n

Andrew Kahng (University of California San Diego)<\/em><\/strong><\/em><\/p>\n\n\n\n

This project aims to improve the efficacy, trust, and security of agentic systems applied to integrated-circuit electronic design automation (IC EDA). The team will develop OpenROAD V2, a cloud-native, AI-assisted extension of the DARPA-funded OpenROAD IC EDA tool. OpenROAD V2 will expose design and code \u201cknobs\u201d in a controlled, declarative way, allowing LLM-guided agents to tune tool behavior without unsafe arbitrary edits. These innovations will enable runtime improvements of 3-10X and design outcome improvements of 15-20%, while providing a trusted \u201csandbox\u201d for AI agents to self-adapt OpenROAD flows to specific technologies and designs. This work will establish a foundation for ML-driven and agentic EDA research, opening new opportunities in chip design research, education and industrial practice.<\/p>\n\n\n\n

\n\n\n\n

Towards Agentic LLM for Chiplet Design<\/strong><\/p>\n\n\n\n

Jun Zhang (Arizona State University)<\/em><\/strong><\/strong><\/p>\n\n\n\n

This project aims to build an agentic AI model for end-to-end electronic design automation (EDA), spanning HDL generation, verification, and physical layout. The framework will build upon Reinforcement Learning from Internal Feedback (RLIF), which learns chip design tasks without the need of external labels. The approach is promising for chip design, a domain with scarce open-source data. By fine-tuning base models with RLIF and integrating verification agents and back-end tasks, the team plans to develop and automate workflows across the design stack with a unified, open-source EDA toolchain. The framework will further enable power-performance optimizations from natural language interfaces. <\/p>\n\n\n\n

<\/div>\n\n\n\n\n\n

<\/h3>\n","protected":false},"excerpt":{"rendered":"

Microsoft is delighted to contribute as a lead supporter of the National AI Research Resource (NAIRR) pilot. (opens in new tab) The mission of the NAIRR pilot aligns with our commitment to broaden AI research and spur innovation by providing greater computing resources to AI researchers and engineers in academia and non-profit sectors. As part […]<\/p>\n","protected":false},"featured_media":1125327,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","footnotes":""},"research-area":[13556,198583,13545,13554,13553,13558,13559],"msr-locale":[268875],"msr-impact-theme":[],"msr-pillar":[],"class_list":["post-1125291","msr-project","type-msr-project","status-publish","has-post-thumbnail","hentry","msr-research-area-artificial-intelligence","msr-research-area-ecology-environment","msr-research-area-human-language-technologies","msr-research-area-human-computer-interaction","msr-research-area-medical-health-genomics","msr-research-area-security-privacy-cryptography","msr-research-area-social-sciences","msr-locale-en_us","msr-archive-status-active"],"msr_project_start":"","related-publications":[],"related-downloads":[],"related-videos":[],"related-groups":[],"related-events":[],"related-opportunities":[],"related-posts":[],"related-articles":[],"tab-content":[],"slides":[],"related-researchers":[{"type":"user_nicename","display_name":"Jessica Young","user_id":43761,"people_section":"Related people","alias":"jessicayoung"}],"msr_research_lab":[],"msr_impact_theme":[],"_links":{"self":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1125291","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project"}],"about":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-project"}],"version-history":[{"count":108,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1125291\/revisions"}],"predecessor-version":[{"id":1163388,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-project\/1125291\/revisions\/1163388"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media\/1125327"}],"wp:attachment":[{"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/media?parent=1125291"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=1125291"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=1125291"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=1125291"},{"taxonomy":"msr-pillar","embeddable":true,"href":"https:\/\/www.microsoft.com\/en-us\/research\/wp-json\/wp\/v2\/msr-pillar?post=1125291"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}