Call for Participation:

Workshop on Emerging Deep Learning Accelerators

Early Registration Deadline
December 25, 2018

1st Workshop on Emerging Deep Learning Accelerators (EDLA)
in conjunction with the HiPEAC 2019 Conference
January 21, 2019
Valencia, Spain

Early Registration Deadline: December 25, 2018

Deep Learning is receiving much attention these days due to remarkable performance achieved in several fields (e.g. Computer Vision, Speech, Translations, etc), although this brings some challenges to hardware architects and computation optimization researchers. Deep Learning models are generally very large in memory and require many computation instructions to train and perform inferences. Accelerating these operations has obvious advantages, first by reducing the energy consumption (e.g. in data centers) and secondly, making these models usable on smaller devices at the edge of the Internet. This workshop on Emerging Deep Learning Accelerators (EDLA) is intended to bring together researchers from academia and industry to discuss requirements, opportunities, challenges and next steps in developing novel approaches for accelerating deep neural networks. The timing of this workshop is ideal, with European regulations tightening data privacy, thus forcing more computations/inferences to be performed at the Edge.


14:00–14:05 Welcome (José Cano, Valentin Radu)

14:05–15:00 Keynote: Milliwatt Human-Quality Speech Recognition (Antonio González, Universitat Politècnica de Catalunya, Barcelona, Spain)

15:00–15:30 Paper presentations

Exploring NEURAghe: A Highly Parameterized APSoC-based CNN Inference Accelerator (15 min)
(Paolo Meloni, Alessandro Capotondi, Deriu Gianfranco, Michele Brian, Francesco Conti, Davide Rossi, Daniela Loi, Marco Carreras, Luigi Raffo and Luca Benini)

Towards Mapping Lift to Deep Neural Network Accelerators (15 min)
(Naums Mogers, Aaron Smith, Dimitrios Vytiniotis, Michel Steuwer, Christophe Dubach and Ryota Tomioka)

15:30–16:00 Coffee break

16:00–17:30 Paper presentations

Energy Efficient Binarized Neural Networks with Adaptive Voltage and Frequency scaling (20 min)
(Jose Nunez-Yanez)

A Streaming Deep Learning Accelerator with Selective Binarization (20 min)
(Sumanta Chaudhuri, Xuecan Yang, Laurence Likforman and Lirida Naviner)

AI Accelerator Latencies in Hybrid Vehicular Simulation (20 min)
(Jussi Hanhirova, Vesa Hirvisalo, Anton Debner and Matias Hyyppä)

Towards efficient mapping of BNNs onto embedded targets using Tensorflow/XLA (15 min)
(Christoph Gratl, Manfred Mücke, Günther Schindler and Holger Fröning)

AI Pipeline – bringing AI to you (15 min)
(Miguel de Prado, Jing Su, Rozenn Dahyot, Rabia Saeed, Lorenzo Keller and Noelia Vallez)

17:30–17:30 Closing remarks (José Cano, Valentin Radu)

José Cano – University of Glasgow
Valentin Radu – University of Edinburgh
David Gregg – Trinity College Dublin
Nuria Pazos – University of Applied Sciences (HES-SO)
Elliot Crowley – University of Edinburgh
Miguel de Prado – ETH Zurich
Jack Turner – University of Edinburgh
Andrew Mundy – ARM Research
Tim Llewellynn – NVISO