From Crash-and-Recover to Sense-and-Adapt: Building Machines with Variability-Resistant Software

Speaker: Rajesh K. Gupta , UC San Diego

Date: Friday, October 25, 2013

Time: 3:00 PM to 4:30 PM Note: all times are in the Eastern Time Zone

Refreshments: 2:45 PM

Public: Yes

Location: 32-D463 - Star Conference Room

Event Type:

Room Description:

Host: Professor Arvind, CSG - CSAIL - MIT

Contact: Sally O. Lee, 3-6837, sally@csail.mit.edu

Relevant URL:

Speaker URL: None

Speaker Photo:
None

Reminders to: seminars@csail.mit.edu

Reminder Subject: TALK: From Crash-and-Recover to Sense-and-Adapt: Building Machines with Variability-Resistant Software

Uncertainty in computing has grown from a performance characterization challenge to a major reliability
headache. Molecular assemblies in nanoscale devices do not behave like chiseled machines with tight
tolerances. Computing machines today are largely ignorant of the variability in the behavior of underlying
components from device to device, chip to chip, their wear over time save for thermal sensing in limited
energy/power constrained applications. Designers sandbag designs with guardbands that threaten to flatten
the very scaling curve that has driven growth of microelectronics over the past decade. This talk explores
an alternate universe where sensing of the ongoing computation, its physical environment provides important
data to adjust software/computation at different levels. In this talk, I will discuss our experiments to
characterize variability, program structuring and task scheduling that can make a software stack robust
against variations in the computing environment.

This talk represents part of the effort at NSF Expeditions in Computing program on Variability.

Bio:
Rajesh K. Gupta is a QUALCOMM professor in Computer Science and Engineering at UC San Diego.
His research focus is on energy efficiency from algorithms, devices to systems that scale from IC chips,
data centers to commercial buildings. His past contributions include SystemC modeling and SPARK
parallelizing high-level synthesis, both of which have been incorporated into industrial practice. Earlier
Gupta lead or co-lead DARPA-sponsored efforts under the Data Intensive Systems (DIS) and Power Aware
Computing and Communications (PACC) programs on role of adaptation in energy efficient system
architectures. His ongoing projects are focused on mitigating microelectronic variability and creating
non-volatile storage/memory systems. In recent years, Gupta and his students have received a best paper
award at IEEE/ACM DCOSS’08 and a best demonstration award at IEEE/ACM IPSN/SPOTS’05. Gupta
received a BTech in EE from IIT Kanpur, MS in EECS from UC Berkeley and a PhD in Electrical Engineering
from Stanford University. Gupta is a Fellow of the IEEE. Gupta serves as chair of Computer Science and as
associate director of the Qualcomm Institute at UCSD.

Research Areas:

Impact Areas:

This event is not part of a series.

Created by Sally O. Lee at Monday, October 21, 2013 at 4:31 PM.