X
Xinyi Zhang
Researcher at University of Pittsburgh
Publications - 18
Citations - 435
Xinyi Zhang is an academic researcher from University of Pittsburgh. The author has contributed to research in topics: Computer science & Speedup. The author has an hindex of 8, co-authored 13 publications receiving 249 citations.
Papers
More filters
Proceedings ArticleDOI
Accuracy vs. Efficiency: Achieving Both through FPGA-Implementation Aware Neural Architecture Search
TL;DR: Field Programmable Gate Arrays (FPGAs) are used as a vehicle to present a novel hardware-aware NAS framework, namely FNAS, which will provide an optimal neural architecture with latency guaranteed to meet the specification and is the very first hardware aware NAS.
Posted Content
Accuracy vs. Efficiency: Achieving Both through FPGA-Implementation Aware Neural Architecture Search.
TL;DR: In this paper, the authors use Field Programmable Gate Arrays (FPGAs) as a vehicle to present a novel hardware-aware NAS framework, namely FNAS, which will provide an optimal neural architecture with latency guaranteed to meet the specification.
Journal ArticleDOI
Achieving Super-Linear Speedup across Multi-FPGA for Real-Time DNN Inference
TL;DR: Super-LIP as mentioned in this paper employs multiple FPGAs to cooperatively run DNNs with the objective of achieving super-linear speed-up against single-FPGA design, which can achieve 3.48× speedup, compared to the state-of-the-art single FPGA design.
Journal ArticleDOI
Achieving Super-Linear Speedup across Multi-FPGA for Real-Time DNN Inference
TL;DR: Super-LIP as mentioned in this paper employs multiple FPGAs to cooperatively run DNNs with the objective of achieving super-linear speed-up against single-FPGA design, which can achieve 3.48x speedup, compared to the state-of-the-art single FPGA design.
Posted Content
DAC-SDC Low Power Object Detection Challenge for UAV Applications
TL;DR: The 55th Design Automation Conference (DAC) held its first System Design Contest (SDC) in 2018 as mentioned in this paper, which attracted more than 110 entries from 12 countries and included 95 categories and 150k images, and the hardware platforms include Nvidia's TX2 and Xilinx's PYNQ Z1.