http://prism.sejong.ac.kr

CALL FOR PAPER

In conjunction with the 41st International Symposium on Computer Architecture(ISCA-41)


 



Workshop Overview

Welcome to the 2nd annual workshop on parallelism in mobile platforms(PRISM-2)!

Mobile platforms are an increasingly important computing environment for many people and businesses. It is both historic and symbolic that PCs were outsold by smartphones in 2011. Rapid advances in wireless network technology like 4G LTE and multicore based mobile architectures continue to fuel the development of exciting new applications on mobile platforms, which, in turn, attract elevated attention to the mobile platforms. While performance and power are two most critical design goals for computing platforms across the board today, mobile platforms are perhaps the most sensitive to the right balancing of the two design goals due to their battery powered nature. Furthermore, mobile platforms are subject to other unique design constraints like: very limited form-factor, no active cooling, and constantly changing external environment. Typically, environment changes are continuously "sensed" for contextualizing a given system. One of the most important principles in designing today's computing systems is to exploit parallelism. Mobile platforms are no exception and we find increasingly more instances of the use of parallelism in them. At the hardware level, there are: multiple processor cores, GPGPU, accelerators, multiple banks of memory, multiple channels to non-volatile memory chips, and multiple radios, to name a few. At the software level, parallel and concurrent threading techniques are commonly employed to improve responsiveness and throughput in the OS and applications alike. We anticipate that future mobile platforms will make more extensive and creative use of parallelism. This workshop focuses on how parallelism is, and can be, utilized in hardware, software and their interaction in order to improve the user experiences with mobile platforms.

 

Area of Interests

One of the most important principles in designing today's computing systems is to exploit parallelism. Mobile platforms are no exception and we find increasingly more instances of the use of parallelism in them. At the hardware level, there are: multiple processor cores, GPGPU, accelerators, multiple banks of memory, multiple channels to non-volatile memory chips, and multiple radios, to name a few. At the software level, parallel and concurrent threading techniques are commonly employed to improve responsiveness and throughput in the OS and applications alike. We anticipate that future mobile platforms will make more extensive and creative use of parallelism.
This workshop focuses on how parallelism is, and can be, utilized in hardware, software and their interaction in order to improve the user experiences with mobile platforms. Topics of particular interest include, but are not limited to:

- Emerging parallel application processor architectures and hardware features in mobile platforms;
- Compelling future applications on mobile platforms that call for unprecedented parallelism;
- Mobile GPGPU architectures and programming models;
- Hardware accelerators for mobile applications;
- Storage architectures in mobile platforms;
- Radio and networking architectures in mobile platforms;
- Compiler support for parallel mobile platforms;
- OS support to accommodate and promote parallelism in mobile platforms;
- Experiences in parallel mobile applications development;
- Novel techniques to improve responsiveness by exploiting parallelism;
- Novel techniques to improve performance/energy by exploiting parallelism;
- Mobile platform performance evaluation methodologies;
- Application benchmarks for mobile platforms;
- Characterization of emerging workloads on mobile platforms; and
- Impact and interaction of emerging technologies to mobile platforms

Keynote

 Speaker
 
- Robert Cohn, Senior Principal Engineer at Intel
- Title : HTML5: A Ubiquitous Runtime for Concurrent Programs
- Astract :HTML started as a way to share documents. Over time, it's capabilities for interactivity improved and with the advent of HTML5, it now has the capability to make full applications. The ubiquity of browsers makes HTML one of the most widely deployed application runtimes. It is possible to deploy the same application on wearable devices like smart watches, mobile devices like phones and tablets, desktop computers, and even automobiles. HTML programs are highly concurrent. Applications use asynchronous IO to send and receive data from remote computers, the UI operates asynchronously from the program logic, and programs are structured as event handlers. HTML applications don't have any of the familiar concurrency concepts like threads, processes, semaphores, locks, or barriers. However, all of the usual concurrency pitfalls still appear: load balancing, races, deadlocks, and livelocks. I will describe the HTML programming model and how it relates to concurrency, the solutions that have been adopted, and the problems that remain.
- Bio : Robert Cohn is a Senior Principal Engineer at Intel where he serves as architect for Intel's XDK tools (http://xdk.intel.com) for developing cross platform applications with HTML5. Previously he led a path finding group that developed tools for parallelization, CPU design, and software analysis. Robert created the Pin instrumentation system. With 1900 citations for its primary publication, Pin is used widely for experimentation across many domains of research as well as in Intel product design and Intel software products. Robert has a PhD. in Computer Science from Carnegie Mellon and a BA from Cornell University.

Invited Talk

  Speaker
 
- Vijay Janapa Reddi, Assistant Professor in the Department of Electrical and Computer Engineering at The University of Texas
- Title : Architecting for the Mobile Web: Where We've Been, Where We're Heading, and What We Need to Address
- Astract : Mobile computing has been around for over a decade. However, the mobile ecosystem has never been as pervasive a computing platform as it is today. An important stimulant in its evolution and growth is the advancement of Web (HTML5, CSS3) and Internet (HTTP 2.0) technologies. With the emergence of these technologies, users have grown to expect rich user experiences that demand high-performance architectures in a mobile form factor. Thus far, the development of such architectures has been driven largely by lessons borrowed from the development of desktop-level systems. However, battery and passive cooling constraints introduce challenges that limit our ability to sustain those development trends. Going forward, we need mobile-specific hardware/software co-design mechanisms that support the needs of both existing and emerging applications. By looking at the past, examining the present, and peeking at the future, in this talk we understand the fundamental challenges, as well as the opportunities, that lay ahead of us for architecting next-generation mobile systems that achieve both high performance and energy efficiency without compromising end-user experiences.
- Bio : Vijay Janapa Reddi is an Assistant Professor in the Department of Electrical and Computer Engineering at The University of Texas in Austin. He received a Ph.D. in Computer Science from Harvard University. His broad research interests are in the area of computer systems, focusing on the interactions between hardware and software to explore new opportunities and synergies for cross-layer solutions that improve processor- and system-level power, performance and reliability. His recent work has focused on mobile systems. He has authored several papers in these areas and has received IEEE Micro Top Picks and Best Paper Awards in Computer Architecture. In the past, he has also worked on architecture and compiler aspects at Intel, VMware, AMD Research and Microsoft Research.

Workshop Program

  8:50 - 9:00 : Welcome and Workshop Introduction

  9:00 - 10:15 : Keynote Speech & Q&A: HTML5: A Ubiquitous Runtime for Concurrent Programs, Robert Cohn, Intel

  10:15 - 10:45 : Break

Paper Session I : Core/Processor Architecture for Mobile Platforms

  10:45 - 11:10 : "Looking into Heterogeneity: When Simple is Faster", Jae Min Kim, Sung Woo Chung, Korea University, Seung Kyu Seo, LG Electronics (paper)

  11:10 - 11:35 : "Spatial computation on a homogeneous, many-core architecture", Daniel Bates, Alex Bradbury, Andreas Koltes, Robert Mullins, University of Cambridge (paper)

  11:35 - 12:00 : "A Case for MVPs: Mixed-Precision Vector Processors", Albert Ou, UC Berkeley, Quan Nguyen, Yunsup Lee, Krste Asanovic, University of California, Berkeley (paper)

  12:00 - 1:30 : Lunch break

  1:30 - 2:30 : Invited Talk & Q&A: Architecting for the Mobile Web: Where We've Been, Where We're Heading, and What We Need to Address, Vijay Janapa Reddi, UT Austin

  2:30 - 3:00 : break

Paper Session II : Software and Runtime Environment Oprimization for Mobile Platforms

  3:00 - 3:25 : "Breaking Cyclic-Multithreading Parallelization with XML Parsing", Simone Campanoni, Svilen Kanev, Kevin Brownell, Gu-Yeon Wei, David Brooks, Harvard University (paper)

  3:25 - 3:50 : "Automatic Runtime Selection of Best Hardware for Data-Parallel JavaScript Kernels via Lifelong Profiling", Younghwan Oh, Channoh Kim, Xianglan Piao, Jae W. Lee, Sungkyunkwan University (paper)

SUBMISSION GUIDELINE

Submit a 2-page presentation abstract to a web-based submission system (through the here) by March 31, 2014. Notification of acceptance will be sent out by April 24, 2014. Final presentation material (to be posted on the workshop web site) due June 4, 2014. For additional information regarding paper submissions, please contact the organizers.

IMPORTANT DATES

Abstract submission : March 31, 2014  April 7, 2014
Author notification : April 24, 2014
Final camera-ready paper : June 4, 2014
Workshop : June 14, 2014

Workshop Organizers

Sangyeun Cho, Samsung Electronics Co. and Univ. of Pittsburgh (cho@cs.pitt.edu)
Hyesoon Kim, Georgia Tech (hyesoon@cc.gatech.edu)
Hsien-Hsin Lee, Georgia Tech (leehs@gatech.edu)
Giho Park, Sejong Univ. (ghpark@sejong.ac.kr)

Program Committee

Murali Annavaram, USC
Seungjae Baek, Dankook Univ.
Jesse Beu, ARM
Calin Cascaval, Qualcomm
Koji Inoue, Kyushu Univ.
Minjang Kim, Qualcomm
Yoonbong Kim, SK Hynix
Masaaki Kondo, Univ. of Tokyo
Scott Mahlke, Univ. of Michigan
Margaret Martonosi, Princeton
Trevor Mudge, Univ. of Michigan
Santosh Pande, Georgia Tech
Vijay Janapa Reddi, UT Austin
Chulho Shin, LG Electronics
Youngmin Shin, Samsung Electronics Co.


Web Chair

Hongyeol Lim, Sejong Univ. (hylim@sju.ac.kr)

Prior PRISM

PRISM-1