T
Tayler Hetherington
Researcher at University of British Columbia
Publications - 14
Citations - 1897
Tayler Hetherington is an academic researcher from University of British Columbia. The author has contributed to research in topics: Deep learning & Efficient energy use. The author has an hindex of 8, co-authored 14 publications receiving 1642 citations.
Papers
More filters
Posted Content
Reduced-Precision Strategies for Bounded Memory in Deep Neural Nets
Patrick Judd,Jorge Albericio,Tayler Hetherington,Tor M. Aamodt,Natalie Enright Jerger,Raquel Urtasun,Andreas Moshovos +6 more
TL;DR: This work investigates how using reduced precision data in Convolutional Neural Networks affects network accuracy during classification and proposes a method for finding a low precision configuration for a network while maintaining high accuracy.
Proceedings ArticleDOI
MemcachedGPU: scaling-up scale-out key-value stores
TL;DR: GNoM, a software framework enabling energy-efficient, latency bandwidth optimized UDP network and application processing on GPUs, is introduced and GNoM handles the data movement and task management to facilitate the development of high-throughput UDP network services on GPUs.
Journal ArticleDOI
Proteus: Exploiting precision variability in deep neural networks
Patrick Judd,Jorge Albericio,Tayler Hetherington,Tor M. Aamodt,Natalie Enright Jerger,Raquel Urtasun,Andreas Moshovos +6 more
TL;DR: Proteus is a layered extension over existing DNN implementations that maintains the native precision of the compute engine by converting to and from a fixed-point reduced precision format used in memory.
Journal ArticleDOI
Value-Based Deep-Learning Acceleration
Andreas Moshovos,Jorge Albericio,Patrick Judd,Alberto Delmas Lascorz,Sayeh Sharify,Tayler Hetherington,Tor M. Aamodt,Natalie Enright Jerger +7 more
TL;DR: This article summarizes the recent work on value-based hardware accelerators for image classification using Deep Convolutional Neural Networks (CNNs) by exploiting runtime value properties that are difficult or impossible to discern in advance.
Proceedings ArticleDOI
EDGE: Event-Driven GPU Execution
TL;DR: An event-driven GPU execution model that enables non-CPU devices to directly launch preconfigured tasks on a GPU without CPU interaction is proposed, and it is estimated that EDGE can reduce the kernel launch latency by 4.4x compared to the baseline CPU-launched approach.