Research Bits

All early computers were "synchronous"---the activities of the various parts of the motherboard were synchronized with a centrally generated clock. That design prevails to this day.

In contrast, CMS fabricated the world's first asynchronous microprocessor in 1989, in Alain Martin's Async VLSI lab. Because of its pioneering clockless design, this computer chip would run even at super-low voltage. The speed of computation simply adjusted itself to the amount of power available. The exceptional robustness of the microprocessor to voltage variations was dramatically demonstrated by running it off the power of a potato: the resulting potato chip ran at 50 kHz at 0.75 volts, while it was designed to run at 15 MHz when operated at 5 volts.

This research area has matured, and many recent processors use partial asynchronous design to save energy. Asynchronous VLSI is becoming increasingly essential as chip technology evolves, to prevent single-event upsets and soft errors. More about this research can be found here.

Critical events-such as earthquakes-happen. Can we automatically and reliably detect these events and take protective measures in the nick of time? The answer is yes, thanks to the large numbers of sensors currently held and managed by ordinary citizens.

Researchers in CMS are tapping into sensors already installed in cell phones and laptops to solve important problems. Cyber-physical networks are already in place all around us, even in remote areas of the world. What is missing is a theory about how to optimally deal with widely distributed sensor networks that involve dynamic, possibly unreliable communications. Current research focuses on developing sense-and-respond systems for earthquake detection in populated areas as well as radiation detection in crowded areas. Did you know your cell phone could do that? This project is led by Mani Chandy and Andreas Krause.

Today's communication and power networks are undoubtedly the most complex pieces of infrastructure that our world has created-and we rely heavily on them. Both are distributed nonlinear feedback control systems of a massive scale. Their development enabled innovations with impacts far beyond communications and energy.

In CMS we believe that the power network is about to undergo a historic architectural transformation akin to the one the telephone network recently experienced. Power distribution, too, will become more sustainable, intelligent, open yet secure, autonomous, and participatory. This transformation will arise through the integration of information technology with the grid, continual market restructuring, the rise of renewable energy technology and distributed generation, demand response, and development of new products such as electric vehicles. Intense societal awareness of energy and climate issues will accelerate the change. Our goal is to develop engineering and economic theories as well as algorithms to guide this historic transformation. This project is led by Steven Low and Mani Chandy.

Did you update your status on Facebook today? Tweet anything lately? Perhaps unbeknownst to you, you may have helped advance science.

Our understanding of the structure of social networks (and other complex networks) has grown dramatically over the last decade. From this understanding has emerged nearly "universal" properties such as small-world properties (low degrees of separation between people) and heavy tailed degree distributions (vast differences in the number of friends of each member). Our research looks at how these properties can be exploited to solve problems that, without such structure, would be intractable. For example, how might these structures make it easier to design distributed routing, find stable matchings, or understand information cascades? This project is led by Adam Wierman and Andreas Krause.

The fundamental principles developed in computer science over the past 30 years have allowed us to develop electronic systems with billions of components and software with millions of lines of code to perform amazingly complex tasks. Researchers at CMS are developing the next generation of computer science principles, this time for programming information-bearing molecules like DNA and RNA to create artificial biomolecular programs of similar complexity.

The biomolecular programs of life itself give inspiration that this is possible, from the low-level operating system controlling cell metabolism to the high-level code for development, the process by which a single cell becomes an entire organism. This research effort aims to create analogous molecular programs using non-living chemistry, in which computing and decision-making are carried out by chemical processes themselves. Through the creation of molecular programming languages, theory for analyzing them, and experiments for validating them, it will establish "molecular programming" as a field within computer science. Molecular programming will enable a yet-to-be imagined array of applications, from chemical circuits that interact with biological molecules to molecular robotics and nanoscale computing. This effort is led by Erik Winfree; see also the Molecular Programming Project.

Another accepted philosophy bites the dust: no longer is "faster" always "better."

Modern systems must trade off traditional performance goals with energy concerns---running faster lowers delays, but increases power usage. While we have theory and models for discussing the computation, communication, and memory demands of algorithms and systems, we have yet to develop theory for discussing their energy efficiency. CMS researchers aim to fill this gap, making computer science greener in the process. This project is led by Adam Wierman and Steven Low.