This three-year, $599,939 National Science Foundation project grant supports the development of a hardware/software codesign framework for dynamically composing disaggregated hardware systems securely. Funded under the Computer and Information Science and Engineering program, the work aims to address system security and performance challenges to make disaggregated computing practical. Key products include a virtualized remote memory architecture, object-level data tracking for efficient tiering and coherence, and mechanisms for creating and enforcing hardware-based trusted execution environments. The University of California, Davis will evaluate solutions using the GEM5 simulation infrastructure, with models contributed back to advance reproducibility and sustainability of this open-source tool. Outputs are expected to enable large-scale machine learning and data analytics on disaggregated systems in a more performant and cost-effective manner.
Mod # | Description | Reason For Modification | Federal Obligation (Click to sort descending) | Date (Click to sort ascending) |
|---|---|---|---|---|
| Not listed | $599.9k | 7/28/22 |