Mohammad Alizadeh and Virginia Williams were both recognized for “outstanding promise” in computer science.

Two EECS faculty members have been awarded Sloan Research Fellowships to support their research in computer science.  

Mohammad Alizadeh and Virginia Vassilevska Williams are among 126 U.S. and Canadian scholars who will receive the two-year, $60,000 fellowships, the Alfred P. Sloan Foundation announced this week.

The fellowships, presented annually since 1955, “seek to stimulate fundamental research by early-career scientists and scholars of outstanding promise,” according to the foundation. They are awarded to researchers in eight scientific fields “in recognition of distinguished performance and a unique potential to make substantial contributions to their field.”

Alizadeh is the TIBCO Career Development Assistant Professor in EECS and a member of MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL). His research focuses on improving the performance, robustness, and ease of management of future networks and cloud computing systems. Previously a principal engineer at Cisco, he joined the EECS faculty in 2015.

Williams, also a CSAIL member, is the Steven and Renee Finn Career Development Associate Professor in EECS. Her research applies combinatorial and graph theoretic tools to various computational domains. She was honored for work done at Stanford University, where she was an assistant professor of computer science before she joined the MIT faculty in January 2017.

Six other MIT researchers received fellowships to further their research in economics, mathematics, and physics. For more details, see the MIT News article. For a complete list of this year’s winners, visit the Sloan Research Fellowships website.

by Anne Trafton, MIT News Office

by Larry Hardesty | MIT News Office

The butt of jokes as little as 10 years ago, automatic speech recognition is now on the verge of becoming people’s chief means of interacting with their principal computing devices.

In anticipation of the age of voice-controlled electronics, MIT researchers have built a low-power chip specialized for automatic speech recognition. Whereas a cellphone running speech-recognition software might require about 1 watt of power, the new chip requires between 0.2 and 10 milliwatts, depending on the number of words it has to recognize.

In a real-world application, that probably translates to a power savings of 90 to 99 percent, which could make voice control practical for relatively simple electronic devices. That includes power-constrained devices that have to harvest energy from their environments or go months between battery charges. Such devices form the technological backbone of what’s called the “internet of things,” or IoT, which refers to the idea that vehicles, appliances, civil-engineering structures, manufacturing equipment, and even livestock will soon have sensors that report information directly to networked servers, aiding with maintenance and the coordination of tasks.

"Speech input will become a natural interface for many wearable applications and intelligent devices,” says Anantha Chandrakasan, the Vannevar Bush Professor of Electrical Engineering and Computer Science at MIT, whose group developed the new chip. “The miniaturization of these devices will require a different interface than touch or keyboard. It will be critical to embed the speech functionality locally to save system energy consumption compared to performing this operation in the cloud."

“I don’t think that we really developed this technology for a particular application,” adds Michael Price, who led the design of the chip as an MIT graduate student in electrical engineering and computer science and now works for chipmaker Analog Devices. “We have tried to put the infrastructure in place to provide better trade-offs to a system designer than they would have had with previous technology, whether it was software or hardware acceleration.”

Price, Chandrakasan, and Jim Glass, a senior research scientist at MIT’s Computer Science and Artificial Intelligence Laboratory, described the new chip in a paper Price presented last week at the International Solid-State Circuits Conference.

The sleeper wakes

Today, the best-performing speech recognizers are, like many other state-of-the-art artificial-intelligence systems, based on neural networks, virtual networks of simple information processors roughly modeled on the human brain. Much of the new chip’s circuitry is concerned with implementing speech-recognition networks as efficiently as possible.

But even the most power-efficient speech recognition system would quickly drain a device’s battery if it ran without interruption. So the chip also includes a simpler “voice activity detection” circuit that monitors ambient noise to determine whether it might be speech. If the answer is yes, the chip fires up the larger, more complex speech-recognition circuit.

In fact, for experimental purposes, the researchers’ chip had three different voice-activity-detection circuits, with different degrees of complexity and, consequently, different power demands. Which circuit is most power efficient depends on context, but in tests simulating a wide range of conditions, the most complex of the three circuits led to the greatest power savings for the system as a whole. Even though it consumed almost three times as much power as the simplest circuit, it generated far fewer false positives; the simpler circuits often chewed through their energy savings by spuriously activating the rest of the chip.

A typical neural network consists of thousands of processing “nodes” capable of only simple computations but densely connected to each other. In the type of network commonly used for voice recognition, the nodes are arranged into layers. Voice data are fed into the bottom layer of the network, whose nodes process and pass them to the nodes of the next layer, whose nodes process and pass them to the next layer, and so on. The output of the top layer indicates the probability that the voice data represents a particular speech sound.

A voice-recognition network is too big to fit in a chip’s onboard memory, which is a problem because going off-chip for data is much more energy intensive than retrieving it from local stores. So the MIT researchers’ design concentrates on minimizing the amount of data that the chip has to retrieve from off-chip memory.

Bandwidth management

A node in the middle of a neural network might receive data from a dozen other nodes and transmit data to another dozen. Each of those two dozen connections has an associated “weight,” a number that indicates how prominently data sent across it should factor into the receiving node’s computations. The first step in minimizing the new chip’s memory bandwidth is to compress the weights associated with each node. The data are decompressed only after they’re brought on-chip.

The chip also exploits the fact that, with speech recognition, wave upon wave of data must pass through the network. The incoming audio signal is split up into 10-millisecond increments, each of which must be evaluated separately. The MIT researchers’ chip brings in a single node of the neural network at a time, but it passes the data from 32 consecutive 10-millisecond increments through it.

If a node has a dozen outputs, then the 32 passes result in 384 output values, which the chip stores locally. Each of those must be coupled with 11 other values when fed to the next layer of nodes, and so on. So the chip ends up requiring a sizable onboard memory circuit for its intermediate computations. But it fetches only one compressed node from off-chip memory at a time, keeping its power requirements low.

“For the next generation of mobile and wearable devices, it is crucial to enable speech recognition at ultralow power consumption,” says Marian Verhelst, a professor of microelectronics at the Catholic University of Leuven in Belgium. “This is because there is a clear trend toward smaller-form-factor devices, such as watches, earbuds, or glasses, requiring a user interface which can no longer rely on touch screen. Speech offers a very natural way to interface with such devices.”

The research was funded through the Qmulus Project, a joint venture between MIT and Quanta Computer, and the chip was prototyped through the Taiwan Semiconductor Manufacturing Company’s University Shuttle Program.

by Larry Hardesty | MIT News Office 

Those sensors will have to operate at very low powers, in order to extend battery life for months or make do with energy harvested from the environment. But that means that they’ll need to draw a wide range of electrical currents. A sensor might, for instance, wake up every so often, take a measurement, and perform a small calculation to see whether that measurement crosses some threshold. Those operations require relatively little current, but occasionally, the sensor might need to transmit an alert to a distant radio receiver. That requires much larger currents.

Generally, power converters, which take an input voltage and convert it to a steady output voltage, are efficient only within a narrow range of currents. But at the International Solid-State Circuits Conference last week, researchers from MIT’s Microsystems Technologies Laboratories (MTL) presented a new power converter that maintains its efficiency at currents ranging from 500 picoamps to 1 milliamp, a span that encompasses a 2,000,000-fold increase.

“Typically, converters have a quiescent power, which is the power that they consume even when they’re not providing any current to the load,” says Arun Paidimarri, who was a postdoc at MTL when the work was done and is now at IBM Research. “So, for example, if the quiescent power is a microamp, then even if the load pulls only a nanoamp, it’s still going to consume a microamp of current. My converter is something that can maintain efficiency over a wide range of currents.”

Paidimarri, who also earned doctoral and master’s degrees from MIT, is first author on the conference paper. He’s joined by his thesis advisor, Anantha Chandrakasan, the Vannevar Bush Professor of Electrical Engineering and Computer Science at MIT.

Packet perspective

The researchers’ converter is a step-down converter, meaning that its output voltage is lower than its input voltage. In particular, it takes input voltages ranging from 1.2 to 3.3 volts and reduces them to between 0.7 and 0.9 volts.

“In the low-power regime, the way these power converters work, it’s not based on a continuous flow of energy,” Paidimarri says. “It’s based on these packets of energy. You have these switches, and an inductor, and a capacitor in the power converter, and you basically turn on and off these switches.”

The control circuitry for the switches includes a circuit that measures the output voltage of the converter. If the output voltage is below some threshold — in this case, 0.9 volts — the controllers throw a switch and release a packet of energy. Then they perform another measurement and, if necessary, release another packet.

If no device is drawing current from the converter, or if the current is going only to a simple, local circuit, the controllers might release between 1 and a couple hundred packets per second. But if the converter is feeding power to a radio, it might need to release a million packets a second.

To accommodate that range of outputs, a typical converter — even a low-power one — will simply perform 1 million voltage measurements a second; on that basis, it will release anywhere from 1 to 1 million packets. Each measurement consumes energy, but for most existing applications, the power drain is negligible. For the internet of things, however, it’s intolerable.

Clocking down

Paidimarri and Chandrakasan’s converter thus features a variable clock, which can run the switch controllers at a wide range of rates. That, however, requires more complex control circuits. The circuit that monitors the converter’s output voltage, for instance, contains an element called a voltage divider, which siphons off a little current from the output for measurement. In a typical converter, the voltage divider is just another element in the circuit path; it is, in effect, always on.

But siphoning current lowers the converter’s efficiency, so in the MIT researchers’ chip, the divider is surrounded by a block of additional circuit elements, which grant access to the divider only for the fraction of a second that a measurement requires. The result is a 50 percent reduction in quiescent power over even the best previously reported experimental low-power, step-down converter and a tenfold expansion of the current-handling range.

“This opens up exciting new opportunities to operate these circuits from new types of energy-harvesting sources, such as body-powered electronics,” Chandrakasan says.

“This work pushes the boundaries of the state of the art in low-power DC-DC converters, how low you can go in terms of the quiescent current, and the efficiencies that you can achieve at these low current levels,” says Yogesh Ramadass, the director of power management research at Texas Instruments’ Kilby Labs. “You don’t want your converter to burn up more than what is being delivered, so it’s essential for the converter to have a very low quiescent power state.”

The work was funded by Shell and Texas Instruments, and the prototype chips were built by the Taiwan Semiconductor Manufacturing Corporation, through its University Shuttle Program.

by Elizabeth Durant | Office of the Dean for Undergraduate Education

Let’s be honest: Even the most disciplined college students don’t pay attention all the time during their classes. The temptations of a hushed conversation with a classmate, a daydream, or any number of digital distractions can be hard to resist.

But one could argue that senior Lilly Chin had an excellent excuse for tuning out during one of her comparative media studies classes last October: She was taking a 10-minute online test to qualify for the popular TV quiz show “Jeopardy!” It was only offered on one specific date, at one specific time — which happened to be during class — so she had little choice. “I was trying not to get caught by the teacher while I was answering the questions,” she later confided to a “Jeopardy!” film crew, barely suppressing a giggle.

In the end, it was worth the risk. Chin went on to become a contestant on the show, made it to the finals, and walked away with the college championship title and the tidy sum of $100,000. Sworn to secrecy since the show’s January taping, Chin was able to savor the victory tonight at a final episode screening held in Room 4-237, where she was cheered on by dozens of friends and other fans from the MIT community.

The making of a “Jeopardy!” champion

Chin was one of thousands of students from schools around the U.S. who applied in October to be contestants on the show. Of those, 250 were invited to in-person auditions in New York City in November, which consisted of a written test, gameplay, and an interview. She learned she’d made the cut — a total of 15 students and one alternate — in December, and the show was taped Jan. 10-11 in Los Angeles.

Chin, an electrical engineering and computer science major with a minor in mechanical engineering, credits part of her success to her curiosity about media, which led her to also minor in comparative media studies. She loves “investigating different forms of media, whether it’s film, video games, or children’s literature — [it’s] the same curiosity which leads me to seek out factoids about these media, and which tend to get asked about more on ‘Jeopardy!’”

A native of Decatur, Georgia, Chin is no stranger to trivia competitions; she participated in quiz bowls from 5th grade through high school. She prepared for “Jeopardy!” in a myriad of ways, like reviewing her old trivia books, reading web comics, listening to Top 40 music, and generally spending lots of time “goofing off on the internet.” She found creative ways to bolster her knowledge of subjects she didn’t know well; to address a weakness in history, she crammed The Cartoon History of the Modern World.

Chin enlisted the help of MIT friends to study and practice her gameplay, including playing Protobowl, a real-time, multi-player quiz bowl application created by her classmate, senior Kevin Kwok. She also sought advice from two MIT connections who had “Jeopardy!” experience: her former graduate resident tutor, Philip Arevalo (who motivated her to apply for the show), and Pranjal Vachaspati ’14.

Preparation aside, Chin also had a few tricks up her sleeve. One of them was buzzer strategy. “I think the game is actually more about buzzer strategy than trivia,” she says. Timing is everything: buzz too soon, before the show’s host Alex Trebek finishes reading the clue, and your buzzer will get locked out for a fraction of a second — enough time for an opponent to buzz in. The key is to time it precisely when Trebek is done speaking.

Her board strategy paid off, too. In the more conventional approach, contestants work their way through one category, moving from lower-value clues at the top of the board to higher-value clues at the bottom. Others, like Chin, prefer to jump between categories and choose clues further down the board. “It’s a bit of a controversial strategy,” she says. But the advantage is that skipping around the board can throw off your opponents and increase your odds of finding the clue with the Daily Double. “The Daily Doubles aren’t evenly distributed,” Chin explains. “People have run stats and found they tend to be in the fourth row or so.”

Being on the show was “surreal,” Chin recalls, smiling broadly. “There was a moment when all the contestants realized that this was actually happening. After the game, everyone’s hands were shaking.” To combat her own nerves, she channeled her experience on the trap shooting team (part of the MIT Sporting Clays Association), in which players shoot moving clay targets with a shotgun. “The coach is always like, ‘Don’t keep track of the score, just take it one shot at a time,’’ she says, “because especially for shooting, any sport, you need to be calm. As soon as you start thinking, ‘Oh no, what ifs’, then your game gets off and you miss everything. So I think that really helped.”

“Nerd pride”

Throughout the course of the two-week tournament, as word spread about her progress, Chin developed quite a following on the MIT campus. “The best part of being on the show has definitely been the great outpouring of support the MIT community has given me,” she says.  “At first, I was a bit embarrassed about being on national television and tried to keep the whole thing under wraps. But soon, I found that the more people that I told, the more I found that people were eager to help and support me.”

President L. Rafael Reif, Chancellor Cynthia Barnhart, and Vice President and Dean for Student Life Suzy Nelson were among those rooting for her. “If the clue is ‘Nerd Pride,’ the answer must be, ‘What was our overwhelming reaction when we learned that Lilly Chin just won College Jeopardy?’” Reif wrote in an email to Chin. “Even better, for this longtime professor: You’re not just MIT, you're EECS! Lilly, I hope you have a moment to savor this terrific achievement.”

“[Provost] Marty Schmidt and I have decided you are the Tom Brady of ‘Jeopardy’ — great job!” Barnhart said in an email to Chin, following Chin’s second-to-last appearance on the show. Nelson wrote Chin after her strong showing in the first week: “I’m so proud of your Jeopardy performance…Plus, love your strategy of finding those Daily Doubles — bold and fearless.”

Chin developed quite a fan base among MIT students, who found creative ways to show their support, throwing screening parties and sharing their favorite moments on social media. One friend, Shi-Ke Xue ’16, created a series of gifs of Chin on the show and shared them on Reddit.

Looking back, and ahead

In retrospect, Chin admits she feels a bit bad about taking the “Jeopardy!” online test during class in October, adding, “That is one of my favorite classes.” Luckily, her professor, T.L. Taylor, who also followed her progress on “Jeopardy!” (and is now privy to Chin’s secret about the test), loves the anecdote. “How very apropos,” she says, “considering it was a class on games and culture!”

Chin, who plans to begin a PhD program in robotics after graduation, says she’ll use the prize money to pay off college loan debt and to travel to a few research conferences around the world on video game studies — a media that continues to pique her boundless curiosity.

Three EECS faculty members, who all joined the Department in January 2017, have been awarded career-development chairs.

Stefanie Mueller was named the X-Consortium Career Development Assistant Professor. Awarded by EECS, this chair promotes interest in human-computer communications. It was established through the generous contribution of the X Window Consortium in 1993, which coordinated the development of the X Window systems. Mueller, whose research involves developing novel hardware and software systems that advance personal fabrication technologies, directs the Human-Computer Interaction (HCI) Engineering Group at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL). She received a PhD in HCI from the Hasso Plattner Institute in 2016. Previously, she received an MS in in IT-systems engineering from Hasso Plattner and a BS in computer science in media from the Harz University of Applied Science.

David Sontag was named the Hermann L. F. von Helmholtz Career Development Assistant Professor in the Institute for Medical Engineering and Science (IMES). He is also an assistant professor in EECS and a PI in CSAIL. Sontag’s research focuses on machine learning and artificial intelligence; at IMES, he leads a research group that aims to use machine learning to transform health care. Previously, he was an assistant professor in computer science and data science at New York University’s Courant Institute of Mathematical Sciences and a postdoctoral researcher at Microsoft Research New England. He received a PhD and an SM in electrical engineering from MIT and a BS in computer science from the University of California, Berkeley.

Virginia Vassilevska Williams was named the Steven G. (1968) and Renee Finn Career Development Associate Professor. The chair, awarded by EECS, was established through the generous contribution of Steven and Renee Finn, to allow the pursuit of new research and development paths, and to make potentially important discoveries through early-stage research. Williams, whose research applies combinatorial and graph theoretic tools to various computational domains, is also a member of CSAIL. Previously, she was an assistant professor of computer science and a research associate at Stanford University. She received a PhD in computer science from Carnegie Mellon University and a bachelor’s in mathematics and engineering & applied science from Caltech. Most recently, she was among 126 U.S. and Canadian researchers (including three from EECS) who received 2017 Sloan Research Fellowships to support their work.

The Optical Society has named Dirk Englund, a faculty member in the MIT Department of Electrical Engineering and Computer Science, as the recipient of the 2017 Adolph Lomb Medal.

The medal, established in 1940 in honor of the society’s founding treasurer, “recognizes noteworthy contributions made to optics at an early career stage,” according to the 101-year-old professional association. In announcing the award, the Optical Society cited Englund’s “pioneering contributions to scalable solid-state quantum memories in nitrogen-vacancy diamond, high-dimensional quantum key distribution, and photonic integrated circuits for quantum communication and computation.”

Englund heads the Quantum Photonics Group and is a PI in MIT’s Research Laboratory of Electronics (RLE) and the Microsystems Technology Laboratory (MTL). His research focuses on quantum technologies based on semiconductor and optical systems.

Before coming to MIT in 2013, Englund was an assistant professor of Electrical Engineering and of Applied Physics at Columbia University and a post-doctoral fellow at Harvard University. He received a BS in physics from Caltech, followed by an MS in electrical engineering in and PhD in applied physics, both from Stanford University.

“In the eight years since he finished his PhD, Dirk has built a blockbuster research group that is pouring out top-flight research results across a broad spectrum of topics in photonics and quantum information science,” said Englund’s faculty colleague Jeffrey Shapiro, Julius A. Stratton Professor in EECS. “In doing so, he has also demonstrated leadership of multidisciplinary research teams that is extraordinary for someone so early in their faculty career.”

MIT NewsMarch 14, 2017

U.S. News and World Report has again placed MIT’s graduate program in engineering at the top of its annual rankings, continuing a trend that began in 1990, when the magazine first ranked such programs.

The MIT Sloan School of Management also placed highly; it shares with Stanford University the No. 4 spot for the best graduate business program.

This year, U.S. News also ranked graduate programs in the social sciences and humanities. The magazine awarded MIT’s graduate program in economics a No. 1 ranking, along with Harvard University, Princeton University, Stanford, the University of California at Berkeley, and Yale University.

Among individual engineering disciplines, MIT placed first in six areas: biomedical/bioengineering (tied with Johns Hopkins University — MIT’s first-ever No. 1 U.S. News ranking in this discipline); chemical engineering; computer engineering; electrical/electronic/communications engineering; materials engineering; and mechanical engineering (tied with Stanford). The Institute placed second in aerospace/aeronautical/astronautical engineering (tied with Georgia Tech) and nuclear engineering.

In the rankings of graduate programs in business, MIT Sloan moved up one step from its No. 5 spot last year. U.S. News awarded a No. 1 ranking to the school’s specialties in information systems and production/operations, and a No. 2 ranking for supply chain/logistics.

U.S. News does not issue annual rankings for all doctoral programs but revisits many every few years. In its new evaluation of programs in the social science and humanities, the magazine gave MIT’s economics program a No. 1 ranking overall and either first- or second-place rankings for all eight economics specialties listed. MIT’s political science and psychology programs also placed among the top 10 in the nation.

In the magazine’s 2014 evaluation of PhD programs in the sciences, five MIT programs earned a No. 1 ranking: biological sciences (tied with Harvard and Stanford); chemistry (tied with Caltech and Berkeley, and with a No. 1 ranking in the specialty of inorganic chemistry); computer science (tied with Carnegie Mellon University, Stanford, and Berkeley); mathematics (tied with Princeton University, and with a No. 1 ranking in the specialty of discrete mathematics and combinations); and physics.

U.S. News bases its rankings of graduate schools of engineering and business on two types of data: reputational surveys of deans and other academic officials, and statistical indicators that measure the quality of a school’s faculty, research, and students. The magazine’s less-frequent rankings of programs in the sciences, social sciences, and humanities are based solely on reputational surveys.

John Tsitsiklis, an MIT professor of electrical engineering and computer science, has been appointed director of the Laboratory for Information and Decision Systems (LIDS), and associate director of the Institute for Data, Systems, and Society, effective April 1. Tsitsiklis has previously served as acting co-director and co-associate director of LIDS, and as a co-director of the Operations Research Center.

"The fields that the faculty, students, and staff in LIDS are working on — network science, communications, data analytics, optimization, control, and more — are experiencing unprecedented growth, and represent some of the most important frontiers in engineering," says Ian A. Waitz, dean of the School of Engineering and the Jerome C. Hunsaker Professor in the Department of Aeronautics and Astronautics. “I am delighted to have John, who is a distinguished scholar and is deeply committed to our educational enterprise, in such a key role.” 

Tsitsiklis replaces Asu Ozdaglar, professor and associate head of the Department of Electrical Engineering and Computer Science, who has served as director of LIDS since 2014. “Asu has been a dedicated and thoughtful leader of LIDS,” Waitz notes. “LIDS remains a world-class research environment and is contributing more than ever to this growing field."

Tsitsiklis's research interests are in the theory and application of optimization in stochastic, dynamic, and distributed systems. He is a coauthor of several noted books on distributed computation, linear optimization, and probability, and he holds seven U.S. patents.

Tsitsiklis studied mathematics and electrical engineering as an undergraduate, and earned his PhD in electrical engineering from MIT in 1984. He was an acting assistant professor of electrical engineering at Stanford University in 1983-84, then came to the Department of Electrical Engineering and Computer Science at MIT, where he is currently the Clarence J. Lebel Professor of Electrical Engineering. He is a member of the National Academy of Engineering and holds an honorary doctorate from the Catholic University of Louvain. 

by Chad Galts | School of Engineering

Read this article on the MIT News website.