Renegotiating Accelerator Abstractions Workshop
1st Annual Workshop
Program
Date: 16th September 2019.
Workshop Program
| Time | Session | Speaker | Title |
|---|---|---|---|
| 1100 - 1230 (25-min sessions) |
Intro + Big Picture | Welcome (5-min) - Tony Nowatzki (UCLA) Intro (10-min) - Matt Horsnell (Arm) |
|
| Sarita Adve (UIUC) | |||
| Vikram Adve (UIUC) | HPVM: A Uniform Interface and Abstraction for Heterogeneous Parallel Systems | ||
| Arrvindh Shriraman (SFU) | |||
| 1230 - 1400 | Lunch Break | ||
| 1400 - 1530 (22-min sessions) |
DSLs, Compilers, & IRs | Thierry Moreau (UoWashington) | The Past, Present and Future of Deep Learning Acceleration Stacks. |
| Riyadh Baghdadi (MIT) | Tiramisu: A Polyhedral Compiler for Dense and Sparse Deep Learning and More | ||
| Jeff Setter (Stanford) | Halide to Hardware: Exploring accelerators using software languages | ||
| Naums Mogers (UoEdinburgh) | Functional Interface for Performance Portability on Parallel Accelerators | ||
| 1530 - 1600 | Break | ||
| 1600 - 1650 (25-min sessions) |
Programmable Accelerators | Aviral Shrivastava (ASU) | dMazeRunner: Accelerate perfectly nested loops on dataflow accelerators /td> </tr> |
| Ramana Radhakrishnan (Arm) | |||
| 1650 - 1730 | Panel | Sarita Adve, Vikram Adve, Arrvindh Shriraman, Riyadh Baghdadi | Panel Moderator - David Weaver (Arm) |