7 papers
FuSeFL: Fully Secure and Scalable Federated Learning
Sahar Ghoflsaz Ghinani, Elaheh Sadredini
Federated Learning (FL) enables collaborative model training without centralizing client data, making it attractive for privacy-sensitive domains. While existing approaches employ…
SAIL: SRAM-Accelerated LLM Inference System with Lookup-Table-based GEMV
Jingyao Zhang, Jaewoo Park, Jongeun Lee +1
Large Language Model (LLM) inference requires substantial computational resources, yet CPU-based inference remains essential for democratizing AI due to the widespread availability…
No One-Size-Fits-All: A Workload-Driven Characterization of Bit-Parallel vs. Bit-Serial Data Layouts for Processing-using-Memory
Jingyao Zhang, Elaheh Sadredini
Processing-in-Memory (PIM) is a promising approach to overcoming the memory-wall bottleneck. However, the PIM community has largely treated its two fundamental data layouts, Bit-Pa…
A Near-Cache Architectural Framework for Cryptographic Computing
Jingyao Zhang, Elaheh Sadredini
Recent advancements in post-quantum cryptographic algorithms have led to their standardization by the National Institute of Standards and Technology (NIST) to safeguard information…
CryptoSRAM: Enabling High-Throughput Cryptography on MCUs via In-SRAM Computing
Jingyao Zhang, Elaheh Sadredini
Secure communication is a critical requirement for Internet of Things (IoT) devices, which are often based on Microcontroller Units (MCUs). Current cryptographic solutions, which r…
Enabling Low-Cost Secure Computing on Untrusted In-Memory Architectures
Sahar Ghoflsaz Ghinani, Jingyao Zhang, Elaheh Sadredini
Modern computing systems are limited in performance by the memory bandwidth available to processors, a problem known as the memory wall. Processing-in-Memory (PIM) promises to subs…