Project Grant 2552193
- The National Science Foundation Division of Information and Intelligent Systems awarded $464,384 to the University of California Santa Cruz on April 1, 2026, under the Computer and Information Science and Engineering program (CFDA 47.070), supporting research on secure and efficient AI infrastructure using Compute Express Link (CXL)-enabled disaggregated shared memory. The project develops secure and scalable abstractions and architectural designs for AI infrastructure that enable large-scale...
- The National Science Foundation Division of Information and Intelligent Systems awarded $384,296 to the University of California, San Diego on August 1, 2026, under the Computer and Information Science and Engineering program (CFDA 47.070) to develop a disaggregated, heterogeneous context mixing plane for scalable generative AI systems. The research addresses inefficiencies in current transformer-based AI architectures by separating memory-bound context mixing computation from compute-bound...
- The National Science Foundation Division of Information and Intelligent Systems awarded the University of California Irvine $500,000 on October 1, 2026, under the Computer and Information Science and Engineering program (CFDA 47.070) to develop a complete system stack enabling automated compute scaling across cloud computing infrastructure using Compute Express Link (CXL) shared memory. The project spans four technical tasks. The first delivers fine-grained sharing of individual memory locations...
- The National Science Foundation Division of Information and Intelligent Systems awarded $424,237 to The Leland Stanford Junior University on September 1, 2026, under the Computer and Information Science and Engineering program (CFDA 47.070) for research on retention-aware memory systems for artificial intelligence computing. The project develops a computing infrastructure that aligns AI application data lifetimes with heterogeneous memory architectures optimized for different retention...
- The National Science Foundation Division of Information and Intelligent Systems awarded North Carolina State University $268,768 on July 15, 2026, under the Computer and Information Science and Engineering program (CFDA 47.070) to develop methods for efficiently processing irregular data on graphics processing units for artificial intelligence applications. The project addresses computational inefficiencies in current AI frameworks when handling irregular data structures common in large language...
- The National Science Foundation (NSF) Division of Information and Intelligent Systems awarded a $600,000 Project Grant to the University of California, San Diego (UCSD) under the Computer and Information Science and Engineering (CFDA 47.070) program. The 3-year grant, effective July 1, 2023, supports research to develop dynamic neural network architectures that can efficiently enable multimodal perception, including vision, audio, and language processing. The research aims to address...
- The National Science Foundation Division of Information and Intelligent Systems awarded the University of South Florida $301,642 on August 1, 2026, under the Computer and Information Science and Engineering program (CFDA 47.070) to develop memory-centric hardware architectures for energy-efficient artificial intelligence inference on edge devices. The project runs through July 31, 2029, and addresses the challenge of deploying AI in power-constrained, compact devices such as wearable health...
- The National Science Foundation Division of Information and Intelligent Systems awarded the University of California Irvine $344,624 on May 1, 2026, under the Computer and Information Science and Engineering CAREER program to develop communication network architecture for ultra-large-scale chiplet-based systems. The project addresses semiconductor industry challenges in designing systems that integrate heterogeneous chiplets to support high-performance compute, neural networks, and large...
- The National Science Foundation Division of Computer and Network Systems awarded the University of California, Davis $386,710 on April 15, 2026, under the Computer and Information Science and Engineering program (CFDA 47.070) to organize and execute SAIL 2026: Summit for AI Institutes Leadership, the fifth annual program-wide convening of the National AI Research Institutes. The conference takes place November 2–4, 2026, in San Diego, California, and brings together leaders and key personnel...
- The National Science Foundation Division of Electrical, Communications and Cyber Systems awarded $270,000 to the University of California Santa Cruz on September 1, 2026, under the Engineering program (CFDA 47.041) to develop analog and digital semantic communication frameworks using pretrained large artificial intelligence models. The project addresses data transmission efficiency in wireless applications including autonomous systems, augmented reality, and collaborative sensing by transmitting...
The National Science Foundation Division of Information and Intelligent Systems awarded the University of California, Davis $600,000 on September 1, 2026, through the Computer and Information Science and Engineering program (CFDA 47.070) for research rethinking the role of CPUs in machine learning systems. The project investigates methods to expand CPU involvement in managing communication, data movement, and memory-intensive operations across artificial intelligence infrastructure, with the goal of making future AI systems more efficient, adaptable, and affordable. Three interconnected research thrusts develop a software-defined transport architecture for remote direct memory access networks with programmable control of communication functions; create a portable communication framework enabling heterogeneous accelerators and network devices to exchange data without vendor-specific software dependencies; and investigate attention-as-a-service for large language model inference by offloading attention computation and key-value cache management from GPUs to distributed CPUs. The research seeks to advance principles for software-defined communication, portable accelerator networking, and resource-efficient inference in large-scale machine learning systems, potentially supporting larger models, improving resource utilization, and reducing deployment costs for advanced AI services. Performance occurs in Davis, California. The period of performance runs from September 1, 2026, through August 31, 2029.
Mod # | Description | ReasonForModification | Federal Obligation | Date |
|---|---|---|---|---|
| Not listed | $600.0k | 7/29/26 |