Research projects are newly defined each year, reflecting the needs of Center members, and organized by University Site. Each project is selected by formal vote of members at the annual workshop. Summaries and formal presentations on the 11 projects underway for CY24 in SHREC are provided below.
- University of Pittsburgh
-
- P1-24: Reconfigurable Systems (neuromorphic algorithm accelerators, high-throughput data processing, mission-critical sensor processing, machine-learning acceleration)
- P2-24: Intelligent Systems (few-shot learning for space, neuromorphic vision and computing, novel processor architectures, deep-learning kernel benchmarks)
- P3-24: Scalable Systems (hyperscale processing, space network protocols, fault-tolerant GPU computing, deep-learning metrics and performance)
- P4-24: Resilient Systems (hybrid flight hardware, mechanical and thermal systems, adaptable Versal computing, emerging RISC-V architectures)
- P5-24: Secure and Efficient Systems (trust assessment, constellation scalability)
- Brigham Young University
-
- B1-24: Fault-Tolerant Techniques for Heterogenous Computing Architectures (Versal ACAP reliability, reliable deep learning, high-performance memory reliability, radiation testing of heterogeneous devices)
- University of Florida
-
- F1-24: Device & Architecture Studies for Compute Cache Systems (device and algorithm studies, oneAPI cross-platform and ModSim, compute cache for edge computing)
- F2-24: Development of Large AI Applications and Systems (SDN-managed tactical network resilience, reinforcement learning for satellite topology reconfiguration, enhancements for integrationAItor)
- Virginia Tech
-
- V1-24: High-Productivity Computing in Heterogeneous Systems (FPGA productivity and performance, GPU productivity and performance, Chapel GPU parallel and distributed programming)
- V2-24: Immersive Space for High-Performance Analytics (immersive data science, immersive sense-making, immersive semantic interaction, interactive LLM for high-performance sense-making)
- V3-24: Scalable Computational Analytics (democratized high-productivity analytics, scalable graph analytics)