Preliminary Program for the 2009 CANDE workshop

The conference opens with a reception at 6:00 PM on the evening of Thursday, 29 October.


Friday, October 30th
Morning Session: Green Technology (Chair: Jan Rabaey, UCB)
------------------------------
-----------

Green Technology
Jan Rabaey -  UC Berkeley


Green Data Centers
    Massoud Pedram -
University of Southern California

Abstract: Datacenters provide the supporting infrastructure for a wide range of economic activities based on digital information. As such, they are extremely important drivers of economic growth. They are also at the center of societal changes enabling new media for cyber-social interactions. However, the continued growth of datacenters is now hindered by their unsustainable and rising energy needs. Apart from datacenter energy consumption and associated costs, corporations and governments are also concerned about the environmental impact of datacenters, in terms of their CO2 footprint. In my talk I will describe a number of techniques for improving the energy efficiency of datacenters ranging from combined power and cooling optimizations and adaptive control algorithms built on a variety of mathematical optimization frameworks to dynamic resource provisioning and request mapping in datacenters of the future.

Design of Sustainable Computing Environments
Subtitle: Driven by the need to reduce the total cost of ownership 

Chandrakant D. Patel, HP Fellow and Director, Sustainable IT Ecosystem Laboratory, HP Labs 

Abstract:  Continued population growth coupled with increased per capita consumption will soon require the creation of a new generation of cities in emerging economies and around the globe. These cities will be driven by a growing desire for on-demand, just-in-time access to critical resources at affordable costs. Coupled with the increased risks that physical growth places on the carrying capacity of the biosphere, we cannot expect to meet the future needs of society simply by extending the existing infrastructures.  

We propose that the necessary transformation be enabled by a global IT ecosystem made up of billions of service-oriented client devices and thousands of data centers. The IT ecosystem, with data centers at its core, and pervasive measurement at the edges, will need to be seamlessly integrated into future communities to enable need-based provisioning of critical resources. Such a transformation of the IT ecosystem necessitates addressing sustainability of the IT ecosystem itself from a cradle-to-cradle perspective – minimizing the available energy needed for extraction, manufacturing, waste mitigation, transportation, operation and reclamation. This lifecycle based engineering and management will result in achieving a reduced Total Cost of Ownership (TCO) of the IT ecosystem that will serve the billions.  

About the author:  Chandrakant Patel is currently an HP Fellow and Director of the Sustainable Information Technology Laboratory at Hewlett Packard Laboratories. Chandrakant has been a pioneer in microprocessor and system thermo-mechanical architectures, management of available energy as a key resource in “smart” data centers, and most recently, application of the IT ecosystem to enable a net positive impact on the environment. Chandrakant enjoys teaching, and has taught computer aided design at Chabot College for 16 years. He also teaches courses in thermal management at U.C. Berkeley Extension, Santa Clara University and San Jose State University.  Chandrakant has been granted over 100 U.S. patents, several pending. An IEEE Fellow, Chandrakant holds a BSME from U.C. Berkeley, MSME from San Jose State University, and is a licensed professional mechanical engineer in the state of California.

Sustainability 2.0 = Energy + Water

Rajesh Gupta, UC San Diego

Abstract: Efficient generation, storage and use of energy is at the center of "Green Tech" today. To the extent, CANDE audience will allow it, this talk is a careful construction that looks at energy use at the scale of a small city, while avoiding the socio-political dimensions of sustainability. Specifically, we examine the issues associated with energy efficient buildings, building archetypes and their energy footprints from large scale monitoring and collection of longitudinal data. Based on our experiments, we attempt to outline what is realistic and possible with the emerging technologies that make combined optimization of energy use in buildings, and its relationship to the Water resources.

About the speaker:  Rajesh Gupta is a professor of computer science and engineering at UC San Diego, and holds the QUALCOMM endowed chair in embedded microsystems. His research interests are in energy efficient systems that have taken turn towards large-scale energy use in recent years. He directs smart buildings/smart grids task force at UC San Diego in his role as associate director for the California Institute for Telecommunications and Information Technology (CalIT2). He is Editor-In-Chief of IEEE Embedded Systems Letters. He is a fellow of the IEEE.



Lunch keynote: Srini Devadas, MIT

The Optimal Research Problem: Definition, Complexity Analysis and Heuristic Solutions

We are all familiar with devising algorithms to find optimal or near-optimal solutions to combinatorial problems.  Whether these solutions will stand the test of time or whether the problems will remain relevant are much harder questions to answer when the research is being conducted.  This talk will define the Optimal Research Problem (ORP), in which a research problem with maximum future impact has to be chosen given limited knowledge and obeying resource constraints.  We attempt to prove a hardness result for ORP, and attempt the development of heuristics for ORP.


Afternoon: Bio/Medical DA (Chair: Sankar Basu, NSF)
-----------------------------------------

Synthetic Biology: A New Application Area for Design Automation Research
    Chris Myers - University of Utah

EDA tools have facilitated the design of ever more complex integrated circuits each year.  Synthetic biology would also benefit from the development of genetic design automation (GDA) tools.  Synthetic biology has the potential to help us produce drugs more economically, metabolize toxic chemicals, and even modify bacteria to hunt and kill tumors.  There are, however, numerous challenges to design the genetic circuits used in these applications. First, existing GDA tools require biologists to design and analyze genetic circuits at the molecular level, roughly equivalent to the layout level for electronic circuits.  Another serious challenge is that genetic circuits are composed of very noisy components making their behavior more asynchronous, analog, and non-deterministic in nature.  New GDA research is necessary to address these challenges.  Interestingly, future electronic circuits may soon also face many of the same challenges which opens up the very intriguing idea that this research may in the future also be utilized to produce more robust and power efficient electronic circuits.  This talk will briefly describe our first steps in the development of iBioSim, a GDA tool that supports higher levels of abstraction.  This talk will also present some of the important theoretical and computational research problems in this area that will need to be addressed.

EE and neural biology - twins separated at birth?
Lou Scheffer, Howard Hughes Medical Institute

There is great potential for ideas developed for EE to help the upcoming science of quantitative neural biology.  Like EE, the most accurate biological models involve differential equations, both linear and non-linear.  Also, like EE, the most accurate simulations involve direct solutions of these equations in a manner similar to SPICE.  Furthermore, like EE, only some of the many variables affect the output - for example, mammalian neurons are give a response when the voltage at  a single point, the trigger spot, exceeds a certain value.  The value at this spot is a complex function of the many inputs of the neurons.  Perhaps, as in digital design, there are simpler ways to figure out the value at this point without solving for all values at all intervening points.  If this can be done (and it is not yet completely clear which portions of the response must be preserved) then perhaps model order reduction techniques from EE could playa part in biology.  As another example, the neuron to neuron transmission is always statistical, with a small number of vesicles released, each with a (relativley) small number of transmitter molecules.  This is recieved by a small number of ion channels that stay open for a (statistical) value of time.   Although some responses invlove large numbers of vessicles, molecules, or synapses, these are the exception rather than the rule.  Thus statistical methods developed for IC timing may be a good starting point for biological calculations.  These are just two examples - many more are possible and probably necessary.

About the speaker:  Lou Scheffer is a traitor to the field of EE, having recently succumbed to the dark side and accepted a position at the Howard Hughes Medical Institute, studying the construction and wiring of the brain.


Computing with Things Small, Wet, and Random: Design Automation for Nanoscale Technologies and Biological Processes
Marc Riedel, Univ. of MN

Abstract: This talk will discuss techniques for synthesizing circuits and biological systems that are characterized by uncertainty in the way that they are wired or that they execute. We adopt a novel view of computation: instead of transforming definite inputs into definite outputs, circuits and biological systems transform probability values into probability values. The computation is random at the level of bits or protein-protein reactions; nonetheless, in the aggregate, it becomes exact and robust, since the accuracy depends only on the statistical distributions. The talk will present novel circuit constructs that are analog in character but based on digital components. Also, it will present biological constructs that are digital in character in the sense that they deliver robust outcomes. We propose a bio-design automation flow in which synthesis first is performed at a conceptual level, in terms of abstract biochemical reactions -- a task analogous to technology-independent logic synthesis in circuit design. Then the results are mapped onto specific biochemical reactions, selected from libraries -- a task analogous to technology mapping in circuit design. Our method targets the universal DNA substrate developed by Erik Winfree's group at Caltech as the experimental chassis. 

Biography: Marc Riedel has been an Assistant Professor of Electrical and Computer Engineering at the University of Minnesota since 2006. He is also a member of the Graduate Faculty in Biomedical Informatics and Computational Biology. He has held positions at Marconi Canada, CAE Electronics, Toshiba, and Fujitsu Research Labs. He received his Ph.D. and his M.Sc. in Electrical Engineering at Caltech and his B.Eng. in Electrical Engineering with a Minor in Mathematics at McGill University. His Ph.D. dissertation titled "Cyclic Combinational Circuits" received the Charles H. Wilts Prize for the best doctoral research in Electrical Engineering at Caltech. His paper "The Synthesis of Cyclic Combinational Circuits" received the Best Paper Award at the Design Automation Conference. He is a recipient of the NSF CAREER Award.
 

Self-Assembly for the More-Than-Moore Era
Babak Parviz, Univ of Washington

Abstract: Two trends dominate the electronics industry today: miniaturization and integration of multiple functions into single microsystems. As further reduction in size of electronic devices by extending the current top-down fabrication techniques becomes exceedingly challenging, and as the need for the integration of an ever increasing number of small components into a system becomes evident, new approaches are needed to maintain the pace of progress in the sector. A candidate approach for building small and complex systems is to use self-assembly. Self-assembly is the process of spontaneous organization of components into a higher structure. This talk reviews some of the recent work at the University of Washington to engineer and use self-assembly across the size scales for building structures and functional devices. Examples covered include hybrid organic-inorganic devices, structures made using genetically engineered peptides, and microsystems produced with self-assembly including a functional contact lens.


Dinner keynote:  The ITRS Roadmap
Andrew Kahng, UCSD

2009 CANDE 5-year Predictions



Saturday, October 31st

Morning. EDA Business/Funding (Chair: Jason Cong)
------------------------------
-----------
Summary of the 2009 NSF-Design Automation Workshop "Electronic Design Automation Past, Present, and Future"
Sankar Basu, Robert Brayton, Jason Cong

Abstract:  Electronic Design Automation (EDA) has been an immensely successful field, helping to manage the exponential increase in our capability to implement integrated circuits that currently incorporate billions of transistors. At the same time, it fostered and used theories in computation and modeling, successfully combining theory and practice. EDA has completely transformed the way that electronic engineers design and manufacture integrated circuits. It was one of the earliest to engage in inter-disciplinary collaboration, where the computer scientists and engineers in EDA successfully collaborated with electrical engineers, physicists, chemists, theoretical computer scientists, applied mathematics and optimization experts, and application domain specialists. This workshop was organized to 1) reflect on the success of EDA to see if its practice can influence other fields of computer science, and if its methodology can be applied to other application domains, and 2) to review the progress made under the National Design Initiative and evaluate what new directions and topics should be added to the Initiative.

This report contains an overview of the EDA area, its funding history, a discussion of some major challenges for the future, related emerging technologies and how EDA experience may help in developing these technologies, educational aspects and challenges, EDA's relation with CS theorists and how this collaboration can be resurrected, and finally in Section 8, a series of recommendations on what might be done to promote EDA and help with the serious challenges it faces in the future. The recommendations are divided into 1) promoting research, 2) supporting educational programs, and 3) encouraging enhanced collaboration with industry.
 
Gaps and Opportunities in EDA
David Kung, IBM Research

Abstract:  As technology and applications evolve, design tools and methodology evolve as well to adapt to the changes. But when disruptive technology or application emerges, tools and methodology will not be able to keep up and gaps are created. Some species of tools will become extinct and new species will rise up to fill the gaps. In this presentation, I will talk about these gaps with emphasis on high performance chip design. Some on-going and potential disruptions I will focus on are the emergence of  manycore chips, the unreliable devices, the lithography challenges beyond 22 nm, the increased level of integration, and the end of CMOS. The impact and demand on CAD tools will be discussed.

EDA Investment
Lucio Lanza, Lanza techVentures

New Rules for Building a Successful EDA Business
           Rajeev Madhavan, Magma Design Automation

Abstract: Over the last 10 years, a lot has changed in the EDA market.  It used to be that with an idea for a slight tool improvement, a small group of software engineers and some venture funding, you could start a company and within a few years sell it to one of the top players in the EDA industry. As designs became more complex, the interdependencies of timing, area, power, turnaround time and other criteria created the need for a more holistic approach. To develop such solutions, more manpower, truly innovative technology and a larger investment are required. Today, large venture capitalists aren't investing in EDA, and the top EDA companies have dramatically scaled back their acquisition strategies. In this presentation, Rajeev Madhavan will share his experiences as a serial entrepreneur and CEO. He'll present case studies on two very different acquisitions and will describe what it takes to build a successful EDA business today.



  Poster session:


Affiliation Supervisor Presenter
Topic

Colorado State U Prof. Sudeep Pasricha Shirish Bahirat Hybrid Nano-photonic-electric On-chip Comm Architecture

Michigan Tech Prof. Hu Shiyan Jia Wang Timing driven buffer insertion for carbon nano-tube interconnect and copper interconnect

U of Michigan Prof. Igor Markov Jarrod Roy Hardware IP protection and anti-piracy

U of Cincinnati Prof. Ranga Vemuri Hao Xu Aggressive Runtime Leakage Control in DSM CMOS

U of Utah Prof. Priyank Kalla Christopher Condrat Logic Synthesis using optical devices

U of Virginia Prof. Mircea R. Stan Adam Cabe Reliability Sensor Distribution using Scan Chain Insertion

Zhenyu Qi MSN: Memory Sensor for NBTI

UC Berkeley Prof. Jaijeet Roychowdhury Chenjie Gu Non-linear projection based model order reduction for circuits and bio-chemical systems

UCLA Prof. Jason Cong Guojie Luo 3D Physical design flow and 3D physical hierarchy exploration

Yi Zhou Parallel Multi-level Analytical Global Placement on GPU

UIUC Prof. Martin Wong Yan Tan Printed Circuit Board Routing Is Calling for DA Help

Hongbo Zhang Process-Aware 1-D Standard Cell Design

Virginia Tech Prof. Patrick Schaumont Abhranil Maiti Hardware enabled methodologies for protection of software IP

Prof. JoAnn Paul Mwaffaq Otoom Defining and designing Multicore workload specific processors

UT Austin Prof. David Pan Duo Ding CAD Optimizations for On-chip Integration of Silicon Nano-photonics

UC Riverside Prof. Sheldon Tan Thom Jefferson Eguia General Behavioral Thermal Modeling and Characterization for Multi-core Microprocessor Design

Penn State U Prof. Xie Yuan Xiangyu Dong System Level Performance, Energy and Area Estimation for PC-RAM Array

Université Bretagne Sud Prof. Philippe Coussy Ghizlane LHAIRECH-LEBRETON Low Power High Level Synthesis for Designing DSP Applications on FPGA


The CANDE workshop closes after lunch on Saturday - ICCAD starts Sunday evening