G
Gu-Yeon Wei
Researcher at Harvard University
Publications - 216
Citations - 9743
Gu-Yeon Wei is an academic researcher from Harvard University. The author has contributed to research in topics: Deep learning & Computer science. The author has an hindex of 46, co-authored 198 publications receiving 8154 citations. Previous affiliations of Gu-Yeon Wei include Stanford University.
Papers
More filters
Proceedings ArticleDOI
A High-Throughput Maximum a posteriori Probability Detector
TL;DR: This paper presents a maximum a posteriori probability (MAP) detector, based on a forward-only algorithm that can achieve high throughputs and can approach performance close to the channel capacity limit.
Posted Content
EdgeBERT: Optimizing On-Chip Inference for Multi-Task NLP.
Thierry Tambe,Coleman Hooper,Lillian Pentecost,En-Yu Yang,Marco Donato,Victor Sanh,Alexander M. Rush,David Brooks,Gu-Yeon Wei +8 more
TL;DR: EdgeBERT is presented, an in-depth and principled algorithm and hardware design methodology to achieve minimal latency and energy consumption on multi-task NLP inference and enables fully on-chip inference acceleration of NLP workloads with 5.2x and 157x lower energy than that of an un-optimized accelerator and CUDA adaptations on an Nvidia Jetson Tegra X2 mobile GPU, respectively.
Proceedings ArticleDOI
A 3mm 2 Programmable Bayesian Inference Accelerator for Unsupervised Machine Perception using Parallel Gibbs Sampling in 16nm
Glenn G. Ko,Yuji Chai,Marco Donato,Paul N. Whatmough,Thierry Tambe,Rob A. Rutenbar,David Brooks,Gu-Yeon Wei +7 more
TL;DR: A 16nm programmable accelerator for unsupervised probabilistic machine perception tasks that performs Bayesian inference on Probabilistic models mapped onto a 2D Markov Random Field, using MCMC is described.
Proceedings ArticleDOI
Evaluation of voltage interpolation to address process variations
TL;DR: In this paper, a post-fabrication tuning knob called voltage interpolation is proposed to mitigate the performance and power overheads of process variation in advanced fabrication technologies, which can be used to mitigate overhead arising from random and correlated within-die process variations.
Proceedings ArticleDOI
Very Low Voltage (VLV) Design
Ramon Bertran,Pradip Bose,David Brooks,Jeff Burns,Alper Buyuktosunoglu,Nandhini Chandramoorthy,Eric Cheng,Martin Cochet,Schuyler Eldridge,Daniel Friedman,Hans M. Jacobson,Rajiv V. Joshi,Subhasish Mitra,Robert K. Montoye,Arun Paidimarri,Pritish R. Parida,Kevin Skadron,Mircea R. Stan,Karthik Swaminathan,Augusto Vega,Swagath Venkataramani,Christos Vezyrtzis,Gu-Yeon Wei,John-David Wellman,Matthew M. Ziegler +24 more
TL;DR: The performance and system reliability constraints that are key impediments to VLV are discussed and the associated trade-offs across power, performance and reliability are helpful in inferring the optimal operational voltage-frequency point.