Invited Talk - July 11, 11.15, ETZ E81

Abstract

In this talk, I would like to first give an overview of a
project called “PASTE: ultra low-power system and software
technologies for computer vision-based portable devices” which is a
collaborative research project funded by Ministry of Science and
Technology in Taiwan. The project envisions a portable and ultra
low-power device with computer vision functionalities and people
can “paste” it to any place and create new IoT/wearable services.
Researchers from the area of ultra-low voltage circuit and
system-on-chip (SoC) design, heterogeneous multi-core processor and
architecture, context-aware power/thermal/software management, and
machine learning-based computer vision algorithms involve in the
project, and we explore key technologies of such a device/SoC
(called PASTE device/SoC) from a cross-layer and system
perspective.

Next, I would like to present our recent studies in the project.
First, future compute-intensive applications such as
computer-vision programs may utilize heterogeneous computing
resources such as CPUs, GPUs, FPGAs, etc on the device. Developers
may want to write heterogeneous computing software once, and then
the software can run on different target devices with only minor
modifications. We propose tools to improve the performance/energy
portability of such heterogeneous computing software on diverse
computing devices. Second, thermal and power issues are critical
for portable devices. To estimate temperature and static power,
previous studies rely on-chip thermal sensors (diodes). However, an
embedded processor equips only a few thermal sensors and may suffer
from considerable spatial temperature variances across the
processor core, and the static power estimation based on
insufficient temperature information introduces errors. We propose
methods to improve the accuracy of static power estimations.

Bio

Shiao-Li (Charles) Tsao earned his PhD degree in engineering
science from National Cheng Kung University, Taiwan in 1999. His
research interests include energy-aware computing, embedded
software and system, and mobile communication and wireless network.
He was a visiting scholar at Bell Labs, Lucent technologies, USA,
in the summer of 1998, a visiting professor at Dept. of Electrical
and Computer Engineering, University of Waterloo, Canada, in the
summer of 2007, Dept. of Computer Science, ETH Zurich, Switzerland,
in the summer of 2010 and 2011, and 2012-2013, and School of
Information Technologies, University of Sydney in 2015. From 1999
to 2003, Dr. Tsao joined Computers and Communications Research Labs
(CCL) of Industrial Technology Research Institute (ITRI) as a
researcher and a section manager. Dr. Tsao is currently a professor
of Dept. of Computer Science and director of Institute of Computer
Science and Engineering of National Chiao Tung University. Prof.
Tsao has published more than 110 international journal and
conference papers, and has held or applied 23 US patents. Prof.
Tsao received the Research Achievement Awards of ITRI in 2000 and
2004, Highly Cited Patent Award of ITRI in 2007, Outstanding
Project Award of Ministry of Economic Affairs (MOEA) in 2003, and
Advanced Technologies Award of MOEA in 2003. He also received the
Young Engineer Award from the Chinese Institute of Electrical
Engineering in 2007, Outstanding Teaching Award of National Chiao
Tung University, K. T. Li Outstanding Young Scholar Award from ACM
Taipei/Taiwan chapter in 2008, and 2013 Award for Excellent
Contributions in Technology Transfer from National Science Council.