CALL FOR PAPERS
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/IoT applications;
- Storage architectures in mobile/IoT platforms;
- Radio and networking architectures in mobile/IoT 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/IoT platform performance evaluation methodologies;
- Application benchmarks for mobile/IoT platforms;
- Characterization of emerging workloads on mobile/IoT platforms;
- IoT Enabling technologies (smart sensors, energy harvesting, sensor networks, etc.);
The workshop aims at providing a forum for researchers, engineers and students from academia and industry to discuss their latest research in designing mobile platforms and systems, to bring their ideas and research problems to the attention of others, and to obtain valuable and instant feedback from fellow researchers.
2:00 - 2:10 :Welcome and Workshop Introduction |
2:10 - 3:00 :Invited talk I - The Design Challenges and Considerations of Sensor Hub
(Chulho Chung, Standing-egg Inc.)
3:00 - 3:40 :Invited talk II - Esperanto: A Language Extension for Unified Internet-of-Things Programming
(Hanjun Kim, Postech)
3:40 - 4:00 :Break
4:00 - 4:30 :Invited talk III - How can we optimize convolutional neural network designs on mobile and
(Sungjoo Yoo, CSE, Seoul National University)
4:30 - 5:00 :Invited talk IV - Energy Harvesting Nonvolatile Processor for Internet of Things
(Yongpan Liu, Tsinghua University)
5:00 - 5:30 :Invited talk V - Live Code Update for IoT Devices in Energy Harvesting Environments
(Wonsun Ahn, University of Pittsburgh)
Invited talk I
- Chulho Chung, Standing-egg Inc.
- Title : The Design Challenges and Considerations of Sensor Hub (slides)
- Abstract :
As the market of IoT device grows, a relative new term "Sensor Hub, Sensor Fusion" is made by the industry. Also there are various functional and performance demands for specific purpose such as health device, environment monitor device, and drones. Therefore, we will discuss about the definition of sensor hub, design consideration and challenges, and an implementation case.
Invited talk II
- Hanjun Kim, Postech
- Title : Esperanto: A Language Extension for Unified Internet-of-Things Programming
- Abstract :
Despite the promising applicability of the Internet of Things (IoT), developing an IoT service is challenging for programmers because programmers should develop multiple sub-programs of each device in the service and explicitly manage their synchronization. This presentation proposes Esperanto, a new language extension that allows programmers to write only one integrated object oriented program for one IoT service without worrying about synchronization across the diverse IoT devices. Given programmer annotations on each object that indicate IoT devices, the Esperanto compiler automatically partitions the integrated program into multiple sub-programs for each IoT device, and inserts communication and synchronization codes. This work implements 4 IoT services with Esperanto, and shows that Esperanto effectively simplifies IoT programming.
Invited talk III
- Sungjoo Yoo, CSE, Seoul National University
- Title : How can we optimize convolutional neural network designs on mobile and embedded
- Abstract :
Artificial neural networks start to be adopted to address real problems including computer vision tasks and speech recognition. For wide-spread adoptions, both energy efficiency and performance need to be improved on mobile and embedded systems (e.g., self-driving cars, robots, smartphones, etc.) as well as servers. Optimized designs of neural networks involve various areas from algorithm, software/hardware design to chip implementation. This talk first introduces the notion of convolution in the contexts of neuroscience and artificial neural network. Then, it explains recent works on optimizing convolutional neural networks, e.g., pruning, low-rank approximation, bit-width optimization, etc. Finally, it addresses issues in multi-threaded software implementations on GPGPU and ASIC designs for high performance/low power artificial neural networks.
Invited talk IV
- Yongpan Liu, Tsinghua University
- Title : Energy Harvesting Nonvolatile Processor for Internet of Things (slides)
- Abstract :
Energy harvesting has been widely investigated as a promising method of providing power for ultra-low-power applications. Such energy sources include solar energy, radio- frequency (RF) radiation, piezoelectricity, thermal gradients, etc. However, the power supplied by these sources is highly unreliable and dependent upon ambient environment factors. Hence, it is necessary to develop specialized systems that are tolerant to this power variation, and also capable of making forward progress on the computation tasks. With the emerging memory technologies, nonvolatile processors are manufactured with nonvolatile registers and can work reliably under the environments with frequency power interrupts, such as energy harvesting and wireless powered applications. This talk first presents the motivation of nonvolatile processor design and defines the metrics for such transient computing systems. Furthermore, we will illustration the nonvolatile processor from circuit, architecture and system levels. Finally, the latest chip design efforts are reported and future research opportunities and challenges for internet of things will be given out.
Invited talk V
- Wonsun Ahn, University of Pittsburgh
- Title : Live Code Update for IoT Devices in Energy Harvesting Environments
- Abstract :
IoT devices are often closely coupled to the physical world and under tight budget, form factor, and power/energy consumption constraints. These IoT devices are powered using harvested energy and show significantly different computation patterns from traditional embedded systems. Studies have shown that supporting effective post-deployment code update is essential for an IoT system to survive coding bugs as well as to adapt to the fast changing physical environment and its computational demands. This talk will explore the potential of live code updates for these IoT devices. By carefully studying the characteristics of such devices, we first identify the drawbacks of traditional image rewrite code update strategies. We then propose a novel in-place code update strategy which effectively eliminates system down time and has minimal system resource demands.
Submit a 2-page presentation abstract to a web-based submission system (https://cmt3.research.microsoft.com/PRISM2016/) by April 17, 2016. Notification of acceptance will be sent out by May 5, 2016. Final presentation material (to be posted on the workshop web site) due June 6, 2016. For additional information regarding paper submissions, please contact the organizers.
Abstract submission :
April 17, 2016  April 24, 2016
Author notification : May 5, 2016
Final camera-ready paper : June 6, 2016
Workshop : June 18, 2016
Sangyeun Cho, Samsung Electronics Co. (email@example.com)
Hyesoon Kim, Georgia Tech (firstname.lastname@example.org)
Hsien-Hsin Lee, Georgia Tech (email@example.com)
Giho Park, Sejong Univ. (firstname.lastname@example.org)
Vijay Janapa Reddi, UT Austin (email@example.com)
Minkwan Kee, Sejong Univ. (firstname.lastname@example.org)
Seungjin Lee, Sejong Univ. (email@example.com)