High-Efficiency Computation in Embedded Systems
Event details
| Date | 03.03.2014 |
| Hour | 15:00 |
| Speaker | Dr. Phillip Stanley-Marbell, Apple Inc., Cupertino |
| Location | |
| Category | Conferences - Seminars |
Computing systems embedded in our environments often operate under timing constraints and must conserve energy usage. Because they typically perform computation on noisy inputs, or generate outputs for perception by humans, high precision in their results is not always necessary.
This talk will address the research challenge of time-, energy-, and precision-efficient computation in embedded systems, with application to in-situ sensor data analytics. The talk will present theoretical results augmented by empirical data, to provide a deeper understanding of the tradeoff between time- and energy-efficiency through the use of parallelism, and how these tradeoffs are influenced by semiconductor-device and system-architectural properties. The insight gained from this analysis will be used to motivate two generations of embedded multiprocessor platforms developed to enable research in tradeoffs between energy-efficiency, performance, and tolerance to imprecision in computation and communication.
Bio: Phillip Stanley-Marbell received his Ph.D. from Carnegie Mellon University in 2007. He was a post-doctoral researcher at TU Eindhoven until 2008, when he joined IBM Research---Zurich as a permanent Research Staff Member. In 2012 he joined Apple to see his research ideas deployed in real-world products. Prior to completing his Ph.D., he held intern and full-time positions at AT&T / Lucent Bell-Labs, Philips Consumer Communications, Lucent's Data Networking Group, and NEC Research Labs.
Dr. Stanley-Marbell is the author of a programming language textbook published by John Wiley & Sons in 2003, and of over thirty scientific publications and seven patents / patent applications. He is a member of the ACM, IEEE, Sigma Xi, USENIX, and the Swiss Mathematical Society. From 2003--2004, he served as the copy editor for the ACM Mobile Computing and Communications Review journal.
His research interests are in architectures for high-efficiency embedded data processing, approximate compute architectures for future device technologies, and domain-specific programming languages for implementing precision-efficient algorithms.
This talk will address the research challenge of time-, energy-, and precision-efficient computation in embedded systems, with application to in-situ sensor data analytics. The talk will present theoretical results augmented by empirical data, to provide a deeper understanding of the tradeoff between time- and energy-efficiency through the use of parallelism, and how these tradeoffs are influenced by semiconductor-device and system-architectural properties. The insight gained from this analysis will be used to motivate two generations of embedded multiprocessor platforms developed to enable research in tradeoffs between energy-efficiency, performance, and tolerance to imprecision in computation and communication.
Bio: Phillip Stanley-Marbell received his Ph.D. from Carnegie Mellon University in 2007. He was a post-doctoral researcher at TU Eindhoven until 2008, when he joined IBM Research---Zurich as a permanent Research Staff Member. In 2012 he joined Apple to see his research ideas deployed in real-world products. Prior to completing his Ph.D., he held intern and full-time positions at AT&T / Lucent Bell-Labs, Philips Consumer Communications, Lucent's Data Networking Group, and NEC Research Labs.
Dr. Stanley-Marbell is the author of a programming language textbook published by John Wiley & Sons in 2003, and of over thirty scientific publications and seven patents / patent applications. He is a member of the ACM, IEEE, Sigma Xi, USENIX, and the Swiss Mathematical Society. From 2003--2004, he served as the copy editor for the ACM Mobile Computing and Communications Review journal.
His research interests are in architectures for high-efficiency embedded data processing, approximate compute architectures for future device technologies, and domain-specific programming languages for implementing precision-efficient algorithms.
Practical information
- General public
- Free
Organizer
- Prof. Yusuf Leblebici
Contact
- Prof. Yusuf Leblebici