Rajeev Balasubramonian

简介

Professor and Associate Director at the Kahlert School of Computing, University of Utah. Research focuses on computer architecture, particularly memory systems for data-intensive workloads, and accelerators for machine learning, genomic analysis, and security primitives.

代表成果

• FlexProf: Flexible, Side-Channel-Free Memory Access
• PATHFINDER: Practical Real-Time Learning for Data Prefetching
• Hyena: Balancing Packing, Reuse, and Rotations for Encrypted Inference
• XCRYPT: Accelerating Lattice-Based Cryptography With Memristor Crossbar Arrays
• CANDLES: Channel-Aware Novel Dataflow-Microarchitecture Co-Design for Low Energy Sparse Neural Network Acceleration
• OrderLight: Lightweight Memory-Ordering Primitive for Efficient Fine-Grained PIM Computations
• Dvé: Improving DRAM Reliability and Performance On-Demand via Coherent Replication
• SpinalFlow: An Architecture and Dataflow Tailored for Spiking Neural Networks
• Compact Leakage-Free Support for Integrity and Reliability
• Wire-Aware Architecture and Dataflow for CNN Accelerators