Finding teammates on and off the field

As a star athlete in high school, Ben Delhees never dreamed he would one day attend the Massachusetts Institute of Technology. Delhees, an Ohio native, was more of a sports fanatic — he competed in everything from football to basketball to baseball. Yet, one cold November morning, he found himself walking across MIT’s football field as its newest recruit.

“At first, I was nervous, because I didn’t really know what to expect,” he says. “That changed after I met all the guys on the football team. Even though I was just a recruit, everybody was so inviting. There was a kind of family aspect to the staff that really drew me in.”

For the next two days of his visit, Delhees shadowed Kevin Lyons, a first-year student and fellow football player at the time. Between walking to Lyons’ classes and staying with his fraternity, Delta Kappa Epsilon, Delhees began to envision himself being part of the MIT community.

“What stood out to me about MIT was how often they preached the importance of working together as a team,” he recalls. “That’s what I’m all about. After that visit, I fell in love with the place and have never looked back.”

Prospective athletes who apply to MIT are subject to the same academically focused admissions process as all other applicants. After being admitted to the Class of 2021, he arrived on campus a month early to train for pre-season.

The trainings were a nonstop combination of workouts and meetings. Despite the exhaustion, the experience allowed Delhees to quickly bond with all 80 members of the team. “Even though it was mentally and physically challenging, I loved pre-season. You spend all day with the same people and that’s how you get so close,” he says. “It made me feel at home before I even started school.”

Getting to know his fellow teammates helped ease Delhees during the transition to campus. The required curriculum forced Delhees to take courses that were particularly challenging and brought the team together in a whole new way. “All the freshman players were going through the same classes, so it was awesome to have a bunch of friends that were sharing my experience,” he says.

Delhees’ commitment to teamwork also extended beyond his football friendships. He chose to create study groups that welcomed students from all backgrounds, athletes and nonathletes alike. “Unlike in high school, where it’s more common to have social barriers, MIT is great at making inclusive communities,” he says. “I really believe in that message. It doesn’t matter whether you’re in chess club or basketball; we’re in this together.”

After taking a variety of courses, Delhees decided to declare a double major in finance and mathematical economics. He believes that the two topics share an unexpected overlap with football. “To approach a market, you have to consider everyone’s background and decision-making process,” he explains. “It’s the same thing in football. You have to know where other people are coming from before you know where you are.”

Delhees was additionally inspired to pursue finance and economics after taking 15.501 (Introduction to Financial and Managerial Accounting) with Professor Joseph Weber. Conversations after class soon led to an immediate connection. Over the next two years, Weber has continued to serve as a mentor for Delhees’ academic interests. Their close friendship also led Delhees to pursue a year-long accounting research project in Weber’s lab.

The project, which required Delhees to summarize dense accounting data, eventually became much more. Delhees and Weber realized the research could be shared to help teach accounting to the general public. The work became the foundation for Blacktip Research, a startup Delhees co-founded with fellow student Matt Beveridge to educate young people on financial literacy.

“My co-founder and I want to teach simple investing principles in an interactive way, so more people my age can choose how to invest their money,” Delhees says. “With everything going on right now with GameStop and Robinhood, people deserve a better way to understand what’s happening. We’re currently working with professors at MIT Sloan School of Management to get our front-end launched. It’s overall been a very rewarding experience.”

In addition to his friendships made through classes and football, Delhees considers memories with his fraternity brothers as the best part of his MIT experience. Delhees is now the house manager of Delta Kappa Epsilon (DKE), the same fraternity that he stayed with during his early days as a recruit. The role has allowed him to advocate his team-oriented values as a priority for the house.

“I didn’t think I would join a fraternity at first. But after that first visit, I realized it felt more like 40 best friends living in a house together,” he says, fondly. “It’s the same type of atmosphere as the football team. Even if you’ve only known someone for a year, you know they’re willing to stick their head out for you.”

While living with 40 brothers can be challenging, Delhees says the group always finds ways to come together. His leadership and belief in shared camaraderie have guided members through all types of challenges, from cleanup to conflict resolution. Delhees remembers a time when all the house furniture had to be removed to meet a cleaning deadline. With music blasting and everyone helping, the stressful chore became a fun social event. “In the end, it took just a few hours to finish. Everyone was more than happy to be there, and I think that’s why we stay so close,” Delhees says. His brothers have also been there for Blacktip Research, providing encouragement and technical help whenever needed.

With graduation quickly approaching, Delhees hopes to find and create more community-focused atmospheres beyond MIT. He will be working for the next two years as an analyst at Stadium Capital, a hedge fund in Connecticut. The role will teach him more about public equity markets and the investment process.

“One of the exciting parts of the job is just the amount of learning I’m going to have to do,” he says. “There’s so many moving parts to markets that we’re never going to fully comprehend them. But by working together to understand each player’s unique circumstances, we might just have a chance.”

de MIT News https://ift.tt/37Wmiwp

3 Questions: Task Force 2021 and the future of MIT education

MIT’s Task Force 2021 and Beyond has been at work for seven months, charged by President L. Rafael Reif with exploring “how MIT might invent a thriving new future” in a post-Covid world. The effort’s Academic Workstream, which looked specifically at the future of the MIT education, was co-chaired by Anantha Chandrakasan, dean of the School of Engineering and Vannevar Bush Professor of Electrical Engineering and Computer Science, and Melissa Nobles, Kenan Sahin Dean of the School of Humanities, Arts, and Social Sciences and professor of political science.

Chandrakasan and Nobles spoke with MIT News about the recent societal changes that are likeliest to impact teaching and learning at MIT, the themes that arose in their group’s conversations, and the changes that might arise if some of their proposals are adopted.

Q: What changes — due to the events of this past year — do you think will have the most significant impact on an MIT education?

Nobles: Needless to say, many developments over the past year have had a profound impact on how we educate our students.

We’ve learned a great deal about teaching remotely — specifically, what works well and what doesn’t work well. We’ve learned what can be accomplished effectively through online education, and what parts of teaching and learning are most productively done in person.

Because of this experience, once we’re able to return to the classroom, I believe we’ll be much more thoughtful about how to use that most valuable of resources — our time together as teachers and students.

Chandrakasan: Events over the past year, both on and off campus, have also served as valuable reminders of how far we still have to go to realize our highest aspirations for diversity, equity, and inclusion. Those events have caused us all to question what more we can do to educate our students on these important topics. We need to help them think and respond to these issues energetically and creatively. In so doing, we will help our graduates to play a vital role in shaping a more inclusive world.

Q: What were themes that emerged in your group’s discussions and in the ideas that your group put forward?

Nobles: There were a number of suggestions for additions to the curriculum. These proposals centered, in part, around teaching on ethics, racial justice, and structural, systemic, and institutional hierarchies. But we also discussed teaching what one might call the “hidden curriculum”: how to deal with the complexities and uncertainties of life, and how to care for mind, body, and relationships.

More broadly, there were themes of clearly articulating MIT’s social responsibilities, and of supporting students, faculty, and staff in fulfilling these responsibilities — through, for example, experiential learning, or possibly the creation of a Social Impact Fund.

Consistent with this is the need to strengthen our pipeline of underrepresented and minority researchers, considering both our hiring processes as well as our need to provide a more supportive, attractive environment once these individuals have found their professional home at MIT.

Chandrakasan: Continuing on the theme of supporting those in our community, there were also a number of suggestions around the support we provide to undergraduates, graduates, and postdocs. We heard clear interest in improving the spectrum of options available to our students and postdocs for advising, mentoring, and professional development.

Additionally, there was ample conversation about how we might most effectively leverage digital tools, reserving in-class time for that which can best be done in person.

Finally, we’ve heard a number of thoughts about how the increased pace of change in today’s world necessitates more lifelong learning. The advances in learning technologies that were introduced to us as a result of Covid-19 might help us make great strides in this area. We’re thinking deeply about the learning opportunities we provide to our graduates, to professionals more broadly, and to underserved communities.

Q: If some of your group’s key ideas are implemented, how would MIT be different for its students in 10 years?

Chandrakasan: It’s clear that there’s broad interest in MIT retaining its academic rigor and its focus on providing an outstanding education in science, technology, and many other areas of scholarship.

However, it is possible that an MIT education may become somewhat more fluid and permeable, with opportunities for our alumni and others to engage with the Institute in lifelong learning. There are many, many qualified people beyond our campus who would like to be able to look to MIT to help them keep pace with changes in science, engineering, and other areas, such as policy or design.

Nobles: At the same time, our group’s discussions underscored the hope that MIT will provide a more holistic education, with yet more focus on nurturing our students in intellect and spirit. We see this as a powerful way to develop graduates who are even better prepared to serve the nation and the world, and to create positive change as they address the world’s greatest challenges.

de MIT News https://ift.tt/37Qpp9j

Engineering the boundary between 2D and 3D materials

In recent years, engineers have found ways to modify the properties of some “two- dimensional” materials, which are just one or a few atoms thick, by stacking two layers together and rotating one slightly in relation to the other. This creates what are known as moiré patterns, where tiny shifts in the alignment of atoms between the two sheets create larger-scale patterns. It also changes the way electrons move through the material, in potentially useful ways.

But for practical applications, such two-dimensional materials must at some point connect with the ordinary world of 3D materials. An international team led by MIT researchers has now come up with a way of imaging what goes on at these interfaces, down to the level of individual atoms, and of correlating the moiré patterns at the 2D-3D boundary with the resulting changes in the material’s properties.

The new findings are described today in the journal Nature Communications, in a paper by MIT graduate students Kate Reidy and Georgios Varnavides, professors of materials science and engineering Frances Ross, Jim LeBeau, and Polina Anikeeva, and five others at MIT, Harvard University, and the University of Victoria in Canada.

Pairs of two-dimensional materials such as graphene or hexagonal boron nitride can exhibit amazing variations in their behavior when the two sheets are just slightly twisted relative to each other. That causes the chicken-wire-like atomic lattices to form moiré patterns, the kinds of odd bands and blobs that sometimes appear when taking a picture of a printed image, or through a window screen. In the case of 2D materials, “it seems like anything, every interesting materials property you can think of, you can somehow modulate or change by twisting the 2D materials with respect to each other,” says Ross, who is the Ellen Swallow Richards Professor at MIT.

While these 2D pairings have attracted scientific attention worldwide, she says, little has been known about what happens where 2D materials meet regular 3D solids. “What got us interested in this topic,” Ross says, was “what happens when a 2D material and a 3D material are put together. Firstly, how do you measure the atomic positions at, and near, the interface? Secondly, what are the differences between a 3D-2D and a 2D-2D interface? And thirdly, how you might control it — is there a way to deliberately design the interfacial structure” to produce desired properties?

Figuring out exactly what happens at such 2D-3D interfaces was a daunting challenge because electron microscopes produce an image of the sample in projection, and they’re limited in their ability to extract depth information needed to analyze details of the interface structure. But the team figured out a set of algorithms that allowed them to extrapolate back from images of the sample, which look somewhat like a set of overlapping shadows, to figure out which configuration of stacked layers would yield that complex “shadow.”

The team made use of two unique transmission electron microscopes at MIT that enable a combination of capabilities that is unrivalled in the world. In one of these instruments, a microscope is connected directly to a fabrication system so that samples can be produced onsite by deposition processes and immediately fed straight into the imaging system. This is one of only a few such facilities worldwide, which use an ultrahigh vacuum system that prevents even the tiniest of impurities from contaminating the sample as the 2D-3D interface is being prepared. The second instrument is a scanning transmission electron microscope located in MIT’s new research facility, MIT.nano. This microscope has outstanding stability for high-resolution imaging, as well as multiple imaging modes for collecting information about the sample.

Unlike stacked 2D materials, whose orientations can be relatively easily changed by simply picking up one layer, twisting it slightly, and placing it down again, the bonds holding 3D materials together are much stronger, so the team had to develop new ways of obtaining aligned layers. To do this, they added the 3D material onto the 2D material in ultrahigh vacuum, choosing growth conditions where the layers self-assembled in a reproducible orientation with specific degrees of twist. “We had to grow a structure that was going to be aligned in a certain way,” Reidy says.

Having grown the materials, they then had to figure out how to reveal the atomic configurations and orientations of the different layers. A scanning transmission electron microscope actually produces more information than is apparent in a flat image; in fact, every point in the image contains details of the paths along which the electrons arrived and departed (the process of diffraction), as well as any energy that the electrons lost in the process. All these data can be separated out so that the information at all points in an image can be used to decode the actual solid structure. This process is only possible for state-of-the-art microscopes, such as that in MIT.nano, which generates a probe of electrons that is unusually narrow and precise.

The researchers used a combination of techniques called 4D STEM and integrated differential phase contrast to achieve that process of extracting the full structure at the interface from the image. Then, Varnavides says, they asked, “Now that we can image the full structure at the interface, what does this mean for our understanding of the properties of this interface?” The researchers showed through modeling that electronic properties are expected to be modified in a way that can only be understood if the full structure of the interface is included in the physical theory. “What we found is that indeed this stacking, the way the atoms are stacked out-of-plane, does modulate the electronic and charge density properties,” he says.

Ross says the findings could help lead to improved kinds of junctions in some microchips, for example. “Every 2D material that’s used in a device has to exist in the 3D world, and so it has to have a junction somehow with three-dimensional materials,” she says. So, with this better understanding of those interfaces, and new ways to study them in action, “we’re in good shape for making structures with desirable properties in a kind of planned rather than ad hoc way.”

“The present work opens a field by itself, allowing the application of this methodology to the growing research line of moiré engineering, highly important in fields such as quantum physics or even in catalysis,” says Jordi Arbiol of the Catalan Institute of Nanoscience and Nanotechnology in Spain, who was not associated with this work.

“The methodology used has the potential to calculate from the acquired local diffraction patterns the modulation of the local electron momentum,” he says, adding

that “the methodology and research shown here has an outstanding future and high interest for the materials science community.”

de MIT News https://ift.tt/3aUXk2g

3 Questions: Devavrat Shah on curbing online misinformation

The specter of “fake news” looms over many facets of modern society. Waves of online misinformation have rocked societal events from the Covid-19 pandemic to U.S. elections. But it doesn’t have to be that way, according to Devavrat Shah, a professor in the Department of Electrical Engineering and Computer Science and the Institute for Data, Systems and Society. Shah researches the recommendation algorithms that generate social media newsfeeds. He has proposed a new approach that could limit the spread of misinformation by emphasizing content generated by a user’s own contacts, rather than whatever happens to be trending globally. As Congress and a new presidential administration mull whether and how to regulate social media, Shah shared his thoughts with MIT News.

Q: How does misinformation spread online, and do social media algorithms accelerate that spread?

A: Misinformation spreads when a lie is repeated. This goes back thousands of years. I was reminded last night as I was reading bedtime stories to my 6-year-old, from the Panchatantra fables:

A brahmin once performed sacred ceremonies for a rich merchant and got a goat in return. He was on his way back carrying the goat on his shoulders when three crooks saw him and decided to trick him into giving the goat to them. One after the other, the three crooks crossed the brahmin’s path and asked him the same question – “O Brahmin, why do you carry a dog on your back?”

The foolish Brahmin thought that he must indeed be carrying a dog if three people have told him so. Without even bothering to look at the animal, he let the goat go.

In some sense, that’s the standard form of radicalization: You just keep hearing something, without question and without alternate viewpoints. Then misinformation becomes the information. That is the primary way information spreads in an incorrect manner. And that’s the problem with the recommendation algorithms, such as those likely to be used by Facebook and Twitter. They often prioritize content that’s gotten a lot of clicks and likes — whether or not it’s true — and mixes it with content from sources that you trust. These algorithms are fundamentally designed to concentrate their attention onto a few viral posts rather than diversify things. So, they are unfortunately facilitating the process of misinformation.

Q: Can this be fixed with better algorithms? Or are more human content moderators necessary?

A: This is doable through algorithms. The problem with human content moderation is that a human or tech company is coming in and dictating what’s right and what’s wrong. And that’s a very reductionist approach. I think Facebook and Twitter can solve this problem without being reductionist or having a heavy-handed approach in deciding what’s right or wrong. Instead, they can avoid this polarization and simply let the networks operate the way the world operates naturally offline, though peer interactions. Online social networks have twisted the flow of information and put the ability to do so in the hands of a few. So, let’s go back to normalcy.

There’s a simple tweak that could make an impact: A measured amount of diversity should be included in the newsfeeds by all these algorithms. Why? Well, think of a time before social media, when we may chat with people in an office or learn news through friends. Although we are still exposed to misinformation, we know who told us that information, and we tend to share it only if we trust that person. So, unless that misinformation comes from many trusted sources, it is rarely widely shared.

There are two key differences online. First, the content that platforms insert is mixed in with content from sources that we trust, making it more likely for us to take that information at face value. Second, misinformation can be easily shared online so that we see it many times and become convinced it is true. Diversity helps to dilute misinformation by exposing us to alternate points of view without abusing our trust. 

Q: How would this work with social media?

A: To do this, the platforms could randomly subsample posts in a way that looks like reality. It’s important that a platform is allowed to algorithmically filter newsfeeds — otherwise there will be too much content to consume. But rather than rely on recommended or promoted content, a feed could pull most of its content, totally at random, from all of my connections on the network. So, content polarization through repeated recommendation wouldn’t happen. And all of this can — and should — be regulated.

One way to make progress toward more natural behavior is by filtering according to a social contract between users and platforms, an idea legal scholars are already working on. As we discussed, the newsfeed of users impacts their behaviors, such as their voting or shopping preferences. In a recent work, we showed that we can use methods from statistics and machine learning to verify whether or not the filtered newsfeed respects the social contract in terms of how it affects user behaviors. As we argue in this work, it turns out that such contracting may not impact the “bottom line” revenue of the platform itself. That is, the platform does not necessarily need to choose between honoring the social contract and generating revenue.

In a sense, other utilities like the telephone service providers are already obeying this kind of contractual arrangement with the "no spam call list" and by respecting whether your phone number is listed publicly or not. By distributing information, social media is also providing a public utility in a sense, and should be regulated as such.

de MIT News https://ift.tt/3kwUsvN

Driving on the cutting edge of autonomous vehicle tech

In October, a modified Dallara-15 Indy Lights race car programmed by MIT Driverless will hit the famed Indianapolis Motor Speedway at speeds of up to 120 miles per hour. The Indy Autonomous Challenge (IAC) is the world’s first head-to-head, high-speed autonomous race. It offers MIT Driverless a chance to grab a piece of the $1.5 million purse while outmaneuvering fellow university innovators on what is arguably the most iconic racecourse.

But the IAC has implications beyond the track. Stakeholders for the event include Sebastian Thrun, a former winner of the DARPA Grand Challenge for autonomous vehicles, and Reilly Brennan, a lecturer at Stanford University’s Center for Automotive Research and a partner at Trucks Venture Capital. The hosts are well aware that, much like the DARPA Grand Challenge, the IAC has the potential to catalyze a new wave of innovation in the private sector.

Formed in 2018 and hosted by the Edgerton Center at MIT, MIT Driverless comprises 50 highly motivated engineers with diverse skill sets. The team is intent on learning by doing, pushing the boundaries of the autonomous driving field. “There is so much strategy involved in multiagent autonomous racing, from reinforcement learning to AI and game theory,” says systems architecture lead and chief engineer Nick Stathas, a graduate student in electrical engineering and computer science (EECS). “What excites us the most is coming up with our own approaches to problems in autonomous driving — we're looking to define state-of the-art solutions.”

In the lead up to the big day, the team has been testing their algorithms at hackathons and competing in a championship series called RoboRace. The series features 12 races hosted over six events covered by livestream. In this format, MIT Driverless and their competitors program and race a sleek electric vehicle dubbed the DEVBot 2.0. Reminiscent of a Tesla Roadster, the DEVBot was designed specifically to explore the relationship between human and machine.

The twist is that RoboRace blends the physical world with a virtual world dubbed the Metaverse. Teams must traverse the track while interacting with an augmented reality replete with virtual obstacles that raise lap times and collectibles that lower them. “Think of it as real-life racing meets Mario Kart,” says Yueyang “Kylie” Ying ’19, a graduate student in EECS who works in the Path Planning division at MIT Driverless.

For this challenge, Ying and her teammates have developed a unique planning algorithm they call Spline Racer, which determines if and when their vehicle needs to deviate from the most expedient course around the track to avoid obstacles or collect rewards. “Spline Racer essentially computes potential paths and then chooses the best one to take based on total time to negotiate the path and total cost or reward from bumping into obstacles or collectibles along that path,” explains Ying.

MIT is home to cutting-edge research that benefits MIT Driverless whenever the checkered flag is waved. Roboticist and Professor Daniela Rus is just one of their trusted advisors. Rus is director of MIT's Computer Science and Artificial Intelligence Laboratory (CSAIL), the associate director of MIT's Quest for Intelligence Core, and director of the Toyota-CSAIL Joint Research Center, which focuses on the advancement of AI research and its applications to intelligent vehicles.

Sertac Karaman of the MIT Department of Aeronautics and Astronautics also serves as an advisor to the team. In addition to pioneering research in controls and robotics theory, Karaman is a co-founder of Optimus Ride, the leading self-driving vehicle technology company developing systems for geo-fenced environments.

“One of the competitive advantages of our team is that by virtue of being at MIT, we have firsthand access to a rich concentration of research expertise that we can apply to our own development,” says team captain Jorge Castillo, a graduate student in the MIT Sloan School of Management.

Consider the connection between the Han Lab at MIT and MIT Driverless. Assistant professor of electrical engineering and computer science Song Han’s work on efficient computing, particularly his innovative algorithms and hardware systems based on his own deep compression technique for machine learning, is a boon for an autonomous racing team looking to make their algorithms run faster.

“Dr. Han is a big fan of MIT Driverless, and he's been extremely helpful,” says Castillo. "We can only put a limited amount of computing in our car,” he explains, "so the faster we can make our algorithms run, the better we will be able to make them and the faster the car will be able to go safely.”

Think of MIT Driverless as an essential pit stop in the autonomous knowledge pipeline that flows between the Institute and industry. Their mission is to become the hub of applied autonomy at MIT, leveraging the research done on campus to help their engineers develop a broad skill set that is applicable beyond just the specific use case of autonomous driving.

“There are labs at MIT working to solve some of the most complex problems in the world,” says Castillo. “At MIT Driverless, we believe it’s vital to have a place that functions as a proving ground for this research while training the engineers that will help re-imagine the future of the tech industry when it comes to autonomous systems and robotics.”

And the MIT Driverless approach to autonomous vehicle racing, particularly as it pertains to architecture and data processing, is similar to the way industry addresses the self-driving problem for streets and highways — which is just one reason why the team has no shortage of industry sponsors who want to get involved. “We have a tight integration between the components that make the car run,” says Stathas. “From a systems perspective, we have well-defined sub-systems that our industry partners appreciate because it aligns with real-world autonomous vehicle development.”

In addition to gaining access to some of the most brilliant young talent in the world, industry partners can boost brand awareness while participating in the emerging sport of autonomous racing. “We've formed tight bonds with industry-leading companies," says Castillo. “Very often, our sponsors are our biggest fans. They also place their trust in us and want to recruit from us because our engineers are well equipped to perform in the real world.”

de MIT News https://ift.tt/2Pdzwyo

Kelly Metcalf Pate to lead the Division of Comparative Medicine

Kelly Metcalf Pate, an assistant professor of veterinary medicine at Johns Hopkins University School of Medicine, will become the new director of MIT’s Division of Comparative Medicine (DCM) on March 1. Metcalf Pate will replace James Fox, who has been the director of DCM for more than four decades.

At Johns Hopkins, Metcalf Pate served as the associate director of academic training for the research training programs for veterinarians in the Department of Molecular and Comparative Pathology. She also launched and directed the Boehringer Ingelheim Veterinary Scholars Program, a summer research program for veterinary students. In her research lab, she studies how platelet cells interact with other immune cells during viral infections such as HIV and cytomegalovirus.

“My core passions lie in teaching and research,” says Metcalf Pate, who will also join the faculty of the Department of Biological Engineering. “I am looking forward to continuing to work with veterinarian trainees and veterinary students, and in continuing to do research that not only answers the unknown questions in platelet immunology but also helps us to figure out better ways to work with animal models.”

The DCM includes more than 175 researchers, animal care and veterinary technicians, veterinarians, and administrative staff, who oversee animal care at MIT. Its mission includes maintaining the health of the animals as well as performing biomedical research and training students in research and veterinary medicine.

“I am thrilled to welcome Kelly as our new DCM director. She brings a strong research program, enviable managerial skills, a collaborative nature, and a lot of ideas for how DCM can provide even more value to our researchers,” says Maria Zuber, MIT’s vice president for research.

Metcalf Pate says she was drawn to MIT by the collaborative spirit that she observed among the DCM personnel. “It was very inspiring to see how motivated by the research and scientific process everyone was, and how much everybody clearly worked together as a team,” she says.

Originally established in 1975 as the Division of Laboratory Animal Medicine, the division was renamed in 1980 to reflect growing involvement in research and teaching. Fox has directed the division since its inception, and under his leadership, the DCM has greatly expanded MIT’s medical research program and added new on-campus facilities for animal care and research.

“It has been a privilege and a never-ending set of opportunities to work with countless faculty, staff, and students over the years, with the common goal of conducting sound science, being ever mindful of the care and welfare of animals being used in fulfilling that stated mission, and fostering that important mandate for the students experiencing an MIT education,” Fox says.

Metcalf Pate, who will be the Dorothy W. Poitras Associate Professor of Biological Engineering at MIT, says she plans to maintain and build on the structure that Fox has established over the past 45 years.

“I am honored to be inheriting such an amazing legacy and team in DCM. Dr. Fox has built something at MIT that is unrivaled in providing excellent care for the animals while working collaboratively with the research teams. It also benefits from strong support from the leadership and the MIT community, for which I am thankful. I hope to preserve and build upon this during my own time as director,” she says.

She noted that researchers in DCM already have extensive collaborations in place with other departments around MIT, and she hopes to further expand these collaborations, as well as increasing DCM’s research and training partnerships with other institutions in the Boston area.

One of her longer-term goals is to explore the possibility of establishing a center to focus on “refinement research” — the practice of looking at how animal models are used to try to maximize both the quality of data and the welfare of the animals.

Joining the MIT community is something of a homecoming for Metcalf Pate, who grew up in Massachusetts and earned her bachelor’s degree from Boston University in 2003. She also holds a PhD from Johns Hopkins University School of Medicine and a DVM from Purdue University College of Veterinary Medicine.

“I am delighted that Kelly Pate will be assuming the position of director of DCM, with a tenured faculty appointment in BE,” Fox says. “Kelly embodies the tradition of veterinarians from Johns Hopkins, in being committed to conducting first-rate peer reviewed research, an enthusiastic mentor and teacher, and fulfilling the critical responsibility of directing an indispensable, centralized campus-wide animal resource program. She will undoubtedly foster, enhance, and continue the tradition of excellence that has become the trademark of DCM.”

de MIT News https://ift.tt/2ZPzDCf

An aggressive market-driven model for US fusion power development

Electricity generated by fusion power plants could play an important role in decarbonizing the U.S. energy sector by mid-century, says a new consensus study report from the National Academies of Sciences, Engineering, and Medicine, which also lays out for the first time a set of technical, economic, and regulatory standards and a timeline for a U.S. fusion pilot plant that would begin producing energy in the 2035-40 time frame.

To achieve this key step toward commercialization, the report calls for an aggressive public-private effort to produce by 2028 a pilot plant design that can, when built, accommodate any of the developmental approaches seeking to realize fusion’s potential as a safe, carbon-free, on-demand energy source.

These include what it calls the “leading fusion concept, a deuterium-tritium fueled tokamak,” like that being pursued at MIT spinout Commonwealth Fusion Systems (CFS) with support from the Institute’s Plasma Science and Fusion Center (PSFC) and Department of Nuclear Science and Engineering. Martin Greenwald, deputy director of the PSFC, notes that “the report can be seen as confirming and validating the vision that motivated the founding of CFS in 2018.” The new report follows and extends a 2018 National Academies study that (while acknowledging the significant scientific and technical challenges still faced by fusion) saw promise in the tokamak approach, called for continued U.S. participation in the international ITER fusion experiment, and suggested a pilot plant effort .

PSFC director and Hitachi America Professor of Engineering Dennis Whyte helped develop the new study as a member of the National Academies’ Committee on the Key Goals and Innovation Needed for a U.S. Fusion Pilot Plant, which also included representatives from other universities, national laboratories, and private companies. It sought out a broad range of expertise from government, academic, and private-sector sources, including U.S. utilities and energy companies.  

“The biggest thing,” says Whyte, is that the diverse group “came to a consensus that fusion is relevant, and that this effort is important.” Driving factors include utility industry commitments to deep cuts in carbon emissions in coming decades, along with a combination of simultaneous synergistic advances in fusion science and technology, application of new resources from areas outside the traditional fusion community, and particularly the rise of interest in private fusion developers like CFS, which collectively have received some $2 billion in funding in recent years.

There has also been a broad pivot by much of the nation’s fusion research community away from a focus on science and toward a mission of practical energy production. This consensus was expressed in a recent report by the Federal Energy Sciences Advisory Committee (FESAC) that urged the nation to “move aggressively toward the deployment of fusion energy, which could substantially power modern society while mitigating climate change,” and suggested development of a pilot plant. The new National Academies study advances the concept with specifics on what a successful pilot plant would look like.

The report’s authors took a marketplace-driven approach to defining the pilot plant’s characteristics, based on discussions with utilities and other energy-sector organizations that would ultimately be the builders, owners, and operators of fusion generating facilities, says Whyte. “Setting those goal posts is very important, laying out the technical, regulatory, and economic performance requirements for the pilot plant,” he explains. “They’re demanding, but they should be, because that’s what’s needed to make fusion viable.”

Those requirements include a total pilot plant cost of less than $5-6 billion and generating capacity of at least 50 megawatts. In addition to proving the ability to create reliable, sustained net energy gain and power production from fusion for steadily increasing periods of time, says the report, the plant must provide “cost certainty to the marketplace in terms of capital cost, construction time, control of radioactive effluents including tritium, the cost of electricity, and the maintenance/operating schedule and cost.”

These results would inform subsequent construction of first-of-a-kind commercial fusion plants in the 2040s, and then broader propagation of fusion energy facilities onto the grid around mid-century, by which time major U.S. utilities have committed to deep reductions in their carbon emissions.

A key near-term factor in achieving these goals is formation of multiple public-private teams to conceptualize and design aspects of the pilot plant over the next seven years. These include improved fusion confinement and control, materials that can withstand the withering temperatures and stresses produced during fusion, methods of extracting fusion-generated heat and harnessing it for generation, and development of a closed fuel cycle. All are technically challenging and also require close attention to cost, manufacturability, maintainability, and other system-level considerations.

Combining resources from national labs, academic institutions, and private industry is a good approach to addressing these tasks, says Martin Greenwald, deputy director of the PSFC and senior research scientist. “Technologies like fusion come to market through the private sector, especially in the U.S., and once you understand that you can see appropriate roles for government labs that can do basic research, universities that are free to work with private industry, and companies that can use their own capital to pick up and commercialize the work.” Private space programs provide an example, he notes, with companies building rockets and using NASA facilities for things like testing and launch.

“The question,” adds Greenwald, “is whether we can collectively gather the resources and investments and execute in a way that meets the pace. We don’t want to be complacent about how audacious this is, but we have to be audacious if we’re going to meet the need.”

Bob Mumgaard, chief executive officer of CFS, says the new report is another indication of fusion’s growing momentum. In addition to the two National Academies studies, growing private investment, and FESAC’s community-driven recommendations, he points to the January enactment of federal appropriations legislation that funded both domestic and international fusion activities, including ongoing participation in ITER.

“For first time in 40 years, the U.S. government has a policy of building a new energy industry, a whole ecosystem,” says Mumgaard. “The legislation sort of pre-authorized many of the things the National Academies report says are good ideas, like the pivot into energy technology, the more-aggressive timeline, and getting regulation sorted out, which is going pretty well, actually — that’s all in the bill. It lays the groundwork for the broad community to take all this to heart and start doing the work. It’s very different from isolated companies doing their own thing, and universities running experiments, and has been very rapid in terms of how these things usually go. We are entering a whole new era for fusion.”

Cecil and Ida Green Professor Emeritus Ernest Moniz, who served as U.S. secretary of energy during the Obama administration, adds that “The academy report alerts the scientific community, the Congress, and the Biden Administration, which is prioritizing climate change risk mitigation, to the incredible progress over the last years towards fusion as a viable energy source — innovation along several technology pathways, supported largely by private capital. Public-private partnerships can help take several of these technologies to demonstrations in this decade, allowing fusion to be a critical enabler of a decarbonized electric grid before mid-century.”

de MIT News https://ift.tt/3qTuW6b

Basic cell health systems wear down in Huntington’s disease, analysis shows

Using an innovative computational approach to analyze vast brain cell gene expression datasets, researchers at MIT and Sorbonne Université have found that Huntington’s disease may progress to advanced stages more because of a degradation of the cells’ health maintenance systems than because of increased damage from the disease pathology itself.

The analysis yielded a trove of specific gene networks governing molecular pathways that disease researchers may now be able to target to better sustain brain cell health amid the devastating neurodegenerative disorder, says co-senior author Myriam Heiman, associate professor in MIT’s Department of Brain and Cognitive Sciences and an investigator at The Picower Institute for Learning and Memory. Christian Neri of the Sorbonne’s Centre National de la Recherche Scientifique is the co-senior and co-corresponding author of the study published in eLife.

“If we can maintain the expression of these compensatory mechanisms, it may be a more effective therapeutic strategy than just trying to affect one gene at a time,” says Heiman, who is also a member of the Broad Institute of MIT and Harvard.

In the study, the team led by co-corresponding author Lucile Megret created a process called “Geomic” to integrate two large sets of data from Heiman’s lab and one more from University of California at Los Angeles researcher William Yang. Each dataset highlighted different aspects of the disease, such as its effect on gene expression over time, how those effects varied by cell type, and the fate of those cells as gene expression varied.

Geomic created plots of the data that mapped differences pertaining to 4,300 genes along dimensions such as mouse age, the extent of Huntington’s-causing mutation, and cell type (certain neurons and astrocytes in a region of the brain called the striatum are especially vulnerable in Huntington’s). The plots took the form of geometric shapes, like crumpled pieces of paper, whose deformations could be computationally compared to identify genes whose expression changed most consequentially amid the disease. The researchers could then look into how abnormal expression of those genes could affect cellular health and function.

Big breakdowns

The Geomic analysis highlighted a clear pattern. Over time, the cells’ responses to the disease pathology — linked to toxic expansions in a protein called Huntingtin — largely continued intact, but certain highly vulnerable cells lost their ability to sustain gene expression needed for some basic systems that sustain cell health and function. These systems initially leapt into action to compensate for the disease but eventually lost steam.

One of the biggest such breakdowns in an especially vulnerable cell type, Drd-1 expressing neurons, was maintaining the health of energy-producing components called mitochondria. Last year, Heiman’s lab published a study in Neuron showing that in some Huntington’s-afflicted neurons, RNA leaks out of mitochondria provoking a misguided and immune response that leads to cell death. The new findings affirm a key role for mitochondrial integrity and implicate key genes such as Ndufb10, whose diminished expression may be undermine the cell’s network of genes supporting the system.

The Geomic approach also highlighted an especially dramatic decline in the Drd-1 neurons and in astrocytes of expression of multiple genes in pathways that govern endosome regulation, an essential process for determining where proteins go and when they are degraded within the cells. Here, too, key genes like Rab8b and Rab7 emerged as culprits within broader gene networks.

The researchers went on to validate some of their top findings by confirming that key alterations of gene expression were also present in post-mortem samples of brain tissue from human Huntington’s patients.

While mitochondrial integrity and endosome regulation are two particularly strong examples, Heiman says, the study lists many others. The Geomic source code and all the data and visualizations it yielded are publicly accessible on a website produced by the authors.

“We’ve created a database of future targets to probe,” Heiman says.

Neri adds: “This database sets a precise basis for studying how to properly reinstate brain cell compensation in Huntington’s disease, and possibly in other neurodegenerative diseases that share common compensatory mechanisms with Huntington’s disease.”

Key among these could be regulators of genetic transcription in these affected pathways, Heiman says.

“One promising future direction is that among the genes that we implicate in these network effects, some of these are transcription factors,” she says. “They may be key targets to bring back the compensatory responses that decline.”

A new way to study disease

While the researchers first applied Geomic’s method of “shape deformation analysis” to Huntington’s disease, it will likely be of equal utility for studying any neurodegenerative disease like Alzheimer’s or Parkinson’s, or even other brain diseases, the authors says.

“This is a new approach to study systems-level changes, rather than just focusing on a particular pathway or a particular gene,” Heiman says. “I think this is a really nice proof of principle and hopefully we can apply this type of methodology to the study of other genomic data from other disease studies.”

In addition to Heiman, Neri, and Megret, the paper’s other authors are Barbara Gris, Satish Nair, Jasmin Cevost, Mary Wertz, Jeff Aaronson, Jim Rosinski, Thomas Vogt, and Hilary Wilkinson.

The Sorbonne Université, the CHDI Foundation, and the National Institutes of Health supported the research. Heiman’s lab is also supported by the JPB Foundation.

de MIT News https://ift.tt/3uvGkaE

Researchers improve efficiency of next-generation solar cell material

Perovskites are a leading candidate for eventually replacing silicon as the material of choice for solar panels. They offer the potential for low-cost, low-temperature manufacturing of ultrathin, lightweight flexible cells, but so far their efficiency at converting sunlight to electricity has lagged behind that of silicon and some other alternatives.

Now, a new approach to the design of perovskite cells has pushed the material to match or exceed the efficiency of today’s typical silicon cell, which generally ranges from 20 to 22 percent, laying the groundwork for further improvements.

By adding a specially treated conductive layer of tin dioxide bonded to the perovskite material, which provides an improved path for the charge carriers in the cell, and by modifying the perovskite formula, researchers have boosted its overall efficiency as a solar cell to 25.2 percent — a near-record for such materials, which eclipses the efficiency of many existing solar panels. (Perovskites still lag significantly in longevity compared to silicon, however, a challenge being worked on by teams around the world.)

The findings are described in a paper in the journal Nature by recent MIT graduate Jason Yoo PhD ’20, professor of chemistry and Lester Wolfe Professor Moungi Bawendi, professor of electrical engineering and computer science and Fariborz Maseeh Professor in Emerging Technology Vladimir Bulović, and 11 others at MIT, in South Korea, and in Georgia.

Perovskites are a broad class of materials defined by the fact that they have a particular kind of molecular arrangement, or lattice, that resembles that of the naturally occurring mineral perovskite. There are vast numbers of possible chemical combinations that can make perovskites, and Yoo explains that these materials have attracted worldwide interest because “at least on paper, they could be made much more cheaply than silicon or gallium arsenide,” one of the other leading contenders. That’s partly because of the much simpler processing and manufacturing processes, which for silicon or gallium arsenide requires sustained heat of over 1,000 degrees Celsius. In contrast, perovskites can be processed at less than 200 C, either in solution or by vapor deposition.

The other major advantage of perovskite over silicon or many other candidate replacements is that it forms extremely thin layers while still efficiently capturing solar energy. “Perovskite cells have the potential to be lightweight compared to silicon, by orders of magnitude,” Bawendi says.

Perovskites have a higher bandgap than silicon, which means they absorb a different part of the light spectrum and thus can complement silicon cells to provide even greater combined efficiencies. But even using only perovskite, Yoo says, “what we’re demonstrating is that even with a single active layer, we can make efficiencies that threaten silicon, and hopefully within punching distance of gallium arsenide. And both of those technologies have been around for much longer than perovskites have.”

One of the keys to the team’s improvement of the material’s efficiency, Bawendi explains, was in the precise engineering of one layer of the sandwich that makes up a perovskite solar cell — the electron transport layer. The perovskite itself is layered with a transparent conductive layer used to carry an electric current from the cell out to where it can be used. However, if the conductive layer is directly attached to the perovskite itself, the electrons and their counterparts, called holes, simply recombine on the spot and no current flows. In the researchers’ design, the perovskite and the conductive layer are separated by an improved type of intermediate layer that can let the electrons through while preventing the recombination.

This middle electron transport layer, and especially the interfaces where it connects to the layers on each side of it, tend to be where inefficiencies occur. By studying these mechanisms and designing a layer, consisting of tin oxide, that more perfectly conforms with those adjacent to it, the researchers were able to greatly reduce the losses.

The method they use is called chemical bath deposition. “It’s like slow cooking in a Crock-Pot,” Bawendi says. With a bath at 90 degrees Celsius, precursor chemicals slowly decompose to form the layer of tin dioxide in place. “The team realized that if we understood the decomposition mechanisms of these precursors, then we’d have a better understanding of how these films form. We were able to find the right window in which the electron transport layer with ideal properties can be synthesized.”

After a series of controlled experiments, they found that different mixtures of intermediate compounds would form, depending on the acidity of the precursor solution. They also identified a sweet spot of precursor compositions that allowed the reaction to produce a much more effective film.

The researchers combined these steps with an optimization of the perovskite layer itself. They used a set of additives to the perovskite recipe to improve its stability, which had been tried before but had an undesired effect on the material’s bandgap, making it a less efficient light absorber. The team found that by adding much smaller amounts of these additives — less than 1 percent — they could still get the beneficial effects without altering the bandgap.

The resulting improvement in efficiency has already driven the material to over 80 percent of the theoretical maximum efficiency that such materials could have, Yoo says.

While these high efficiencies were demonstrated in tiny lab-scale devices, Bawendi says that “the kind of insights we provide in this paper, and some of the tricks we provide, could potentially be applied to the methods that people are now developing for large-scale, manufacturable perovskite cells, and therefore boost those efficiencies.”

In pursuing the research further, there are two important avenues, he says: to continue pushing the limits on better efficiency, and to focus on increasing the material’s long-term stability, which currently is measured in months, compared to decades for silicon cells. But for some purposes, Bawendi points out, longevity may not be so essential. Many electronic devices such as cellphones, for example, tend to be replaced within a few years anyway, so there may be some useful applications even for relatively short-lived solar cells.

“I don’t think we’re there yet with these cells, even for these kind of shorter-term applications,” he says. “But people are getting close, so combining our ideas in this paper with ideas that other people have with increasing stability could lead to something really interesting.”

Robert Hoye, a lecturer in materials at Imperial College London, who was not part of the study, says, “This is excellent work by an international team.” He adds, “This could lead to greater reproducibility and the excellent device efficiencies achieved in the lab translating to commercialized modules. In terms of scientific milestones, not only do they achieve an efficiency that was the certified record for perovskite solar cells for much of last year, they also achieve open-circuit voltages up to 97 percent of the radiative limit. This is an astonishing achievement for solar cells grown from solution.”

The team included researchers at the Korea Research Institute of Chemical Technology, the Korea Advanced Institute of Science and Technology, the Ulsan National Institute of Science and Technology, and Georgia Tech. The work was supported by MIT’s Institute for Soldier Nanotechnology, NASA, the Italian company Eni SpA through the MIT Energy Initiative, the National Research Foundation of Korea, and the National Research Council of Science and Technology.

de MIT News https://ift.tt/3uqWdPD

Q&A: Ceasar McDowell on better public conversation

Last year leaders in Poughkeepsie, New York, started the Children’s Cabinet, an organization aimed at bolstering cradle-to-career services and support for the city’s kids. With a wide-ranging agenda, they wanted to figure out how to reach the community for input — and worked with Ceasar McDowell, professor of the practice of civic design and associate head of MIT’s Department of Urban Studies and Planning (DUSP). McDowell is an expert in designing public conversations and leads We Who Engage MIT, a project aimed at enhancing outreach.

Now the group has issued a report, titled “The Civic Design Framework: Principles for public conversations during a time of crisis,” detailing the methods they recommend for this — for Poughkeepsie or any organization seeking effective public dialogue. McDowell emphasizes that We Who Engage is a collaborative effort involving students, alumni, and himself, and that the group gives much “ownership” to people to work their own way. MIT News spoke with McDowell about the new report.

Q: What was the genesis of the Civic Design Framework project?

A: This has come out of my work for many years, around the problem we’re facing in this country. We’re living among the most demographically complex set of people who have ever lived together, and we have neither the infrastructure nor the process for democracy to work. Our public engagement process was built on a foundation of exclusion. Going back to the New England town hall and other institutions, the design worked because it was clear who didn’t belong. If you have a group of men who lived together and owned land, their conversations could be structured to get results. But we’re in a different world now, and we haven’t taken on what that means for public engagement. So we have to build something from the ground up again.

Q: This report emphasizes different types of conversations, depending on whether the goal of the dialogue is to reflect, plan, or act, for instance. You say when a group is reflecting, the conversation might be about “framing” or “ideating,” and when it’s planning, it might be “prioritizing” or “deciding.” When a group is acting, the conversation might focus on “implementing” or “monitoring.” Why do those distinctions matter?

A: We talk a lot about the public’s mistrust of institutions, but we don’t talk enough about the people inside institutions mistrusting the public and its ability to make judgements. The only way to change that is to invite the public into the first step, framing what the issue is — to understand how people experience a problem.

But when you’re framing something, you are in a different space cognitively, versus when you’re in the space of imagination or making value choices about a set of policy options. So you have to design conversations to optimize what you’re trying to get out of them.

Designing from the margins is another one of our anchor principles. People who are living at the margins are living with the failures of the [social and political] system. It centers their voice, so you have a better opportunity to find solutions further down the road that are going to work for a broader set of people.

Q: Who do you want to read this report? Community groups? Public officials? Both?

A: If you look at mayoral offices and [other] government offices, we have co-mingled the political process with the engagement process. There are offices set up to do public outreach, but those things are often deeply tied to party politics. They’re also tied to making sure we solve problems quickly — and I’m not saying this is a bad thing — and demonstrating that we can get things done. But the public needs more than just quickly getting things done. The public needs ways of understanding and building connections.

When Barack Obama won the presidency for the first time, they set up an office for public engagement. If you remember that campaign, one of the things that was so powerful was the way it cut across so many different sectors of America. [But] what was never on the table was [using] his contact list to build public engagement and dialogue [after he was in office]. The thing he ignited could not be used to build a deeper democracy, even though it would have been a perfect launchpad for it. We need something that is foremost about the public and its relationship with itself.

Q: But don’t we want public conversation to reach the political sphere?

A: Yes. Absolutely. We want people to live their political lives, but in a way so that political identity is not the only thing they’re considering. We’re all complex. Look at any issue and you realize values aren’t absolute, people are always making tradeoffs. But we don’t have a space to understand each other in the midst of all these tradeoffs.

Q: So in one sense you’re asking us to step back and take a more fundamental look at ourselves, policy goals aside?

A: Part of what’s going on, no matter where you look, is that people are really asking one question: What’s the future of America, and where am I in that future?

We have had “question campaigns” examining that [run through an initiative McDowell helped found, America’s Path Forward]. We ask people: What’s your question about the future of America? We find people who have similar questions and invite them into conversations [in a room] together. But the conversations are not about solving that question. Our focus is: What experiences in your life have gotten you to that question? And when they show up in these rooms, they find people they didn’t expect to be there. And they walk away with a broader understanding of the things they care about. All of this is about getting the public into the kinds of conversations they need to be in.

de MIT News https://ift.tt/2NVZ42c

Data transfer system connects silicon chips with a hair’s-width cable

Researchers have developed a data transfer system that can transmit information 10 times faster than a USB. The new link pairs high-frequency silicon chips with a polymer cable as thin a strand of hair. The system may one day boost energy efficiency in data centers and lighten the loads of electronics-rich spacecraft.

The research was presented at this month’s IEEE International Solid-State Circuits Conference. The lead author is Jack Holloway ’03, MNG ’04, who completed his PhD in MIT’s Department of Electrical Engineering and Computer Science (EECS) last fall and currently works for Raytheon. Co-authors include Ruonan Han, associate professor and Holloway’s PhD adviser in EECS, and Georgios Dogiamis, a senior researcher at Intel.

The need for snappy data exchange is clear, especially in an era of remote work. “There’s an explosion in the amount of information being shared between computer chips — cloud computing, the internet, big data. And a lot of this happens over conventional copper wire,” says Holloway. But copper wires, like those found in USB or HDMI cables, are power-hungry — especially when dealing with heavy data loads. “There’s a fundamental tradeoff between the amount of energy burned and the rate of information exchanged.” Despite a growing demand for fast data transmission (beyond 100 gigabits per second) through conduits longer than a meter, Holloway says the typical solution has been “increasingly bulky and costly” copper cables.

One alternative to copper wire is fiber-optic cable, though that has its own problems. Whereas copper wires use electrical signaling, fiber-optics use photons. That allows fiber-optics to transmit data quickly and with little energy dissipation. But silicon computer chips generally don’t play well with photons, making interconnections between fiber-optic cables and computers a challenge. “There’s currently no way to efficiently generate, amplify, or detect photons in silicon,” says Holloway. “There are all kinds of expensive and complex integration schemes, but from an economics perspective, it’s not a great solution.” So, the researchers developed their own.

The team’s new link draws on benefits of both copper and fiber optic conduits, while ditching their drawbacks. “It’s a great example of a complementary solution,” says Dogiamis. Their conduit is made of plastic polymer, so it’s lighter and potentially cheaper to manufacture than traditional copper cables. But when the polymer link is operated with sub-terahertz electromagnetic signals, it’s far more energy-efficient than copper in transmitting a high data load. The new link’s efficiency rivals that of fiber-optic, but has a key advantage: “It’s compatible directly with silicon chips, without any special manufacturing,” says Holloway.

The team engineered such low-cost chips to pair with the polymer conduit. Typically, silicon chips struggle to operate at sub-terahertz frequencies. Yet the team’s new chips generate those high-frequency signals with enough power to transmit data directly into the conduit. That clean connection from the silicon chips to the conduit means the overall system can be manufactured with standard, cost-effective methods, the researchers say.

The new link also beats out copper and fiber optic in terms of size. “The cross-sectional area of our cable is 0.4 millimeters by a quarter millimeter,” says Han. “So, it’s super tiny, like a strand of hair.” Despite its slim size, it can carry a hefty load of data, since it sends signals over three different parallel channels, separated by frequency. The link’s total bandwidth is 105 gigabits per second, nearly an order of magnitude faster than a copper-based USB cable. Dogiamis says the cable could “address the bandwidth challenges as we see this megatrend toward more and more data.”

In future work, Han hopes to make the polymer conduits even faster by bundling them together. “Then the data rate will be off the charts,” he says. “It could be one terabit per second, still at low cost.”

The researchers suggest “data-dense” applications, like server farms, could be early adopters of the new links, since they could dramatically cut data centers’ high energy demands. The link could also be a key solution for the aerospace and automotive industries, which place a premium on small, light devices. And one day, the link could replace the consumer electronic cables in homes and offices, thanks to the link’s simplicity and speed. “It’s far less costly than [copper or fiber optic] approaches, with significantly wider bandwidth and lower loss than conventional copper solutions,” says Holloway. “So, high fives all round.”

This research was funded, in part, by Intel, Raytheon, the Naval Research Laboratory, and the Office of Naval Research.

de MIT News https://ift.tt/2NVnFEB

Fight or flight? Why individuals react as they do

Why do some people fight and others flee when confronting violence? “This question has been bothering me for quite some time,” says Aidan Milliff, a fifth-year doctoral student who entered political science to explore the strategic choices people make in perilous times.

“We’ve learned a great deal how economic status, identity, and pressure from community shape decisions people make while under threat,” says Milliff. Early in his studies, he took particular interest in scholarship linking economic deprivation to engagement in conflict.

“But I became frustrated by this idea, because even among the poorest of the poor, way more people sit out conflict instead of engaging,” he says. “I thought there must be something else going on to explain why people decide to take enormous risks.”

A window on this problem suddenly opened for Milliff with class 17.S950 (Emotions and Politics), taught by Roger Petersen, the Arthur and Ruth Sloan Professor of Political Science. “The course revealed the cognitive processes and emotional experiences that influence how individuals make decisions in the midst of violent conflict,” he says. “It was extremely formative in the kinds of research I started to do.”

With this lens, Milliff began investigating questions anew, leveraging unusual data sources and novel qualitative and quantitative methods. His doctoral research is yielding fresh perspectives on how civilians experience threats of violence, and, Milliff believes, “providing policy-relevant insights, explaining how individual action contributes to phenomena like conflict escalation and refugee flows.”

First-person accounts

At the heart of Milliff’s dissertation project, “Seeking Safety: The Cognitive and Social Foundations of Behavior During Violence,” are connected episodes of violence in India: an urban pogrom in Delhi in which nearly 3,000 Sikhs died at the hands of Hindus, sparked by the 1984 assassination of Indira Gandhi by her Sikh bodyguards; and the bloody, decade-long separatist civil war by Sikh extremists in Punjab that began in the 1980s.

In search of first-person testimony to illuminate people’s fight-or-flight choices, Milliff lucked out: He located taped oral histories for a large population of Sikhs who had experienced violence in the 1980s. “In these 500 taped histories, people described at a granular level whether they organized to defend their neighborhoods, hid in houses, left the city temporarily or permanently, or tried to pass as Hindu.” He also pursued field interviews in California and India, but didn’t get as far as he’d hoped: “I arrived in India last March, and was there for two weeks of an intended three-month stay when I had to return due to the pandemic.”

This setback did not deter Milliff, who managed to convert the oral histories into text and video data that he’s already begun to plumb, with the help of natural language processing to code people’s decision-making processes. Among his preliminary findings: “People typically appraise their situations in terms of their sense of control and of predictability,” he says.

“When people feel they have a high degree of control but feel that violence is unpredictable, they are more likely to fight back, and when they sense they have neither control nor predictability, and more easily imagine being victims, they flee.”

A Chicago launchpad

Milliff drew inspiration for his doctoral research directly from an earlier graduate project in Chicago with the families of homicide victims.

“I wanted to learn whether people who become angry in response to violence are more likely to seek retribution,” he says. After taping 90 hours of interviews with 31 people, primarily mothers, Milliff shifted his focus. “My initial assumption that everyone would get angry was wrong,” he says. “I found that when people suffer these losses, they might get sad instead, or become fearful.” In unsolved homicides, family members have no perpetrator to target, but instead turn their anger at government that’s let them down, or worry for the safety of surviving family members.

From this project, Milliff took away a crucial insight: “People respond differently to their tragedies, even when their experiences look similar on paper.”

Political violence and its consequences seized Milliff’s interest early on. For his University of Chicago master’s thesis, he sought to understand how many long-running, brutal independence movements fizzle out. “I came away from this program believing that I’d enjoy the day-to-day work of being a professional political scientist,” he says.

Two research experiences propelled him toward that goal. While in college, Milliff assisted in the National Science Foundation-sponsored General Social Survey, a national social survey headquartered in Chicago, where he learned “how a big quantitative data collection exercise works,” he says. Following graduation, a fellowship at the Carnegie Endowment for International Peace immersed him in South Asian military conflict and Indian domestic politics. “I really enjoyed working on these issues and became greatly interested in focusing on the political situation there,” he says.

Attracted by MIT’s security studies community, especially its commitment to research with real-world impact, Milliff came to Cambridge, Massachusetts, primed to delve deeper into the subject of political violence. He first had to navigate the graduate program’s thorough quantitative sequence. “I came to MIT without having taken math after calculus, and I honestly feel fortunate I ended up somewhere that takes the classroom portion of training seriously,” he says. “It has given me new tools I didn’t even know existed.”

These tools are integral to Milliff’s analysis of his singular datasets, and provide the quantitative foundation for informing his policy ideas. If, as his work suggests, people in crisis make decisions based on their sense of control and predictability, perhaps community institutions could bolster citizens’ abilities to imagine concrete options. “Lack of predictability and a sense of control encourage people to make choices that are destabilizing, such as fleeing their homes, or joining a fight.”

Milliff continues to analyze data, test hypotheses, and write up his research, taking time out for biking and nature photography. “When I was headed to graduate school, I decided to take up a hobby that I could do for 15 minutes at a time, something I could do between problem sets,” he says.

While he acknowledges research can be taxing, he takes delight in the moments of discovery and validation: “You spend a lot of time coming up with ideas of how the world works, diving into a pit to see if an idea is right,” he says. “Sometimes when you surface, you see that you might have come up with a possible new way to describe the world.”

de MIT News https://ift.tt/3uqmu0I

Improving sanitation for the world’s most vulnerable people

Last year, women visiting a neonatal clinic at a hospital in Kiboga, Uganda, began using two waterless, standalone bathrooms that offered a cleaner and more private alternative to the pit latrines that are standard in the region.

The deployment was the culmination of years of work by the startup change:WATER Labs, co-founded by two MIT research scientists — and its success showed the company’s potential to improve lives far beyond Uganda.

Half of the world’s population lacks access to toilets that safely manage human waste, with women and children bearing the brunt of the consequences. A child dies every 15 seconds from water-related diseases like diarrhea and cholera. Women and girls without private bathrooms close to their homes are susceptible to high rates of sexual violence. Globally, 45 percent of schools lack proper bathroom facilities, one reason 20 percent of girls drop out of school by the time they start menstruating.

The small but determined team behind change:WATER Labs believes the solution to these problems is an inexpensive, no-flush toilet that can be dropped into any location and work without external power. The toilet, which the company calls the iThrone, uses a proprietary material to evaporate the water content of human waste, shrinking waste by 95 percent, thus preventing waste discharge and simplifying waste collection.

The breathable material takes advantage of the natural tendency of water molecules to move from areas of high moisture to areas of low moisture. CEO and co-founder Diana Yousef, who is also a research affiliate at MIT, says the iThrone allows for waste collection once or twice a month as opposed to every day, transforming the economics of distributed sanitation in low-resource communities.

“We’re essentially turning human waste into clean molecular water, and what’s left over gets collected much less frequently at much lower cost,” Yousef says. “The solution helps customers managing sanitation to be much more scalable, much more sustainable, and much more profitable.”

Today change:WATER Labs has promising early trial results to go along with support from a host of companies, NGOs, and governments. But back when the company was nothing more than an idea, MIT played a pivotal role in making the iThrone concept a reality.

A unique partnership

The seed of change:WATER Labs was planted for Yousef while working on a water treatment initiative with NASA in 2009. Although the project explored ways to recycle water for space agriculture, Yousef wondered if any of the approaches could be used to improve water sustainability in the developing world.

Five years later, she finally put the idea to paper, pitching an early version at MIT’s Water Innovation Prize and the MIT IDEAS Social Innovation Challenge. The experience helped her connect with others at MIT who were interested in the idea, including co-founder Huda Elasaad, a research affiliate in MIT’s D-Lab. Yousef, who earned her undergraduate degree at Harvard University, a PhD at Cornell University, and MBA and MIA degrees from Columbia University, eventually received seed funding to explore the idea from IDEAS and the MIT PKG Center. The support allowed her team to gain access to lab facilities for early testing.

“[The early support from MIT] was a game-changer for us, because you start to have doubts about whether what you’re doing is possible, and when some other entity like MIT takes a bet on you, you start to believe it yourself,” says Yousef, who notes she didn’t have the resources to pursue the idea on her own and was working on a prototype in her kitchen.

MIT’s relationship with the company has continued to evolve in the years since that early bet. MIT’s Environment, Health, and Safety (EHS) Office has helped the startup develop its waste treatment system, and the company benefits from its association with MIT D-Lab, where it collaborates with MIT students from diverse backgrounds.

“We’ve been so very lucky to find such support and collaborators at MIT,” Yousef says. “MIT provides a truly unique ecosystem that cultivates partnerships between innovators within and around MIT to catalyze world-changing innovations. Our breakthrough wouldn’t have been possible without the support from D-Lab, EHS, the PKG Center, and our other partners at MIT.”

On a mission for change

Change:WATER Labs’ toilets were used by about 400 people per week in Uganda before the project was cut short by Covid-19. Yousef says the iThrones proved safe, with minimal odor and no leakage, showing they could be placed close to densely populated areas.

“We have the potential to put safe, hygienic, clean toilets in places that are crowded and close to where people are, and that’s been one of the challenges with other solutions, like composting toilets and others, that don’t fit in crowded communities,” she says.

The toilets also reduced daily waste volumes so much that they were able to operate for weeks at a time without being serviced. Overall, Yousef says feedback from users was overwhelmingly positive as the iThrones provided a safer, cleaner alternative to pit latrines located on the top of a remote hill.

Although travel restrictions have put other iThrone pilots on hold, change:WATER Labs has received funding from the Bill and Melinda Gates Foundation, the United Nations Development Program, and the Turkish government to install its toilets in refugee communities in Turkey later this year.

Private companies have also expressed interest, including two large construction contractors looking to put iThrones in low-income homes in Central America, and two Indian companies seeking to put iThrones in port-a-potties and on transportation and maritime equipment.

Yousef says that inbound interest is indicative of the large global need and pent-up demand for better sanitation options.

“We need new solutions that contain and eliminate human waste while also reducing the amount of water that gets consumed, preventing pollution,” Yousef says. “We solve all of that.”

Yousef says the company never would have reached this point without the MIT community, which she commends for embracing her effort even though she is not an alumna.

“MIT’s willingness to open up its community to the innovators around it allows for things to happen that really don’t happen anywhere else,” she says. “It’s special to be here and it’s really amplified what we’re trying to do.”

de MIT News https://ift.tt/3dEy5mz

Amy Jin named 2021 Gates Cambridge Scholar

MIT senior Amy Jin has won the prestigious Gates Cambridge Scholarship, which offers students an opportunity to pursue graduate study in the field of their choice at the University of Cambridge in the UK. Jin will join the other 23 U.S. citizens in being members of the 20th class of scholars.

Jin, from Pleasanton, California, is completing double majors in biological engineering and electrical engineering and computer science. During her time volunteering as an aquarist at the New England Aquarium, she became fascinated by the power and cruelty of evolution. She wishes to harness the power she observed in natural phenomena, such as the “virgin birth” she witnessed of a baby green anaconda, to inspire engineering designs related to human health. At Cambridge, she plans to conduct research in bioelectronics to develop medical technology for treating neurological disorders.

Her studies at Cambridge will build upon her undergraduate research in the Langer and Traverso labs, where she has worked on evaluating polymer hydrogel drug delivery systems and designing microbiome-based therapies that target inflamed parts of the gastrointestinal tract. In fall 2019, she began work in the Bathe Lab, which uses DNA and RNA to engineer new nanoscale materials for therapeutics and computing. Jin specifically worked on DNA storage and computation. Professor Mark Bathe says, “I was thrilled to learn of the wonderful news that Amy received the Gates Cambridge Scholarship. Amy combines the highest levels of creativity, technical and intellectual depth and breadth, leadership, and motivation for her work to impact society — I cannot imagine a more deserving candidate for this prestigious award.” She is also a NEET Living Machines scholar.

Jin has been an active volunteer at the New England Aquarium and a garden chair for UA Sustainability. She was a member of Terrascope, a first-year learning community that tackles a global issue related to sustainability. She has volunteered with the Petey Greene Program, providing tutoring to incarcerated people. She participated in MIT’s service center Criminal Justice Immersion Program, which brought students to local detention centers, trials courts, and the Massachusetts attorney general’s office. She worked for the Massachusetts Trial Court as a recidivism analyst. She worked as an extern with the Massachusetts Eye and Ear Glaucoma Clinic and has been dedicated to the HMS Family Van, which performs free health screenings in their mobile clinic. She is an Emerson piano scholar and participates in the MIT Chamber Music Society.

Jin was advised in her application by Kim Benard of the Distinguished Fellowships team in Career Advising and Professional Development, who remarks, “Amy represents the best of MIT, pursuing scientific research with passion and excitement. She applies the lessons she has learned volunteering with the aquarium to create innovative solutions for human health. But even more than her abilities in science, she has demonstrated remarkable empathy and compassion for others, demonstrated through her volunteer activities.”

Established by the Bill and Melinda Gates Foundation in 2000, the Gates Cambridge Scholarship provides full funding for talented students from outside the United Kingdom to pursue postgraduate study in any subject at Cambridge University. Since the program’s inception in 2001, there have been 32 Gates Cambridge scholars from MIT.

de MIT News https://ift.tt/37G0NQu

Study: Covid-19 communications featuring racially diverse physicians can improve health outcomes for communities of color

In a new study by MIT professors Esther Duflo, Ben Olken, and Abhijit Banerjee, and physicians Marcella Alsan and Fatima Cody Stanford, along with other doctors and economists, public health video messages featuring a racially diverse set of physicians were found to decrease knowledge gaps about Covid-19 symptoms and transmission — generating important lessons about how we communicate about, and work to mitigate the effects of, the virus. Study results also indicated that, for Black individuals, watching a video with a racially concordant (Black) physician increased the degree to which they sought out more information about Covid-19.

Black and Latinx communities have significantly higher infection and mortality rates and are more likely to have severe symptoms and be hospitalized as a result of the Covid-19 pandemic compared to white communities. Although these disparities stem from many complex factors, including inequalities in access to health care, systemic racism, and the overrepresentation of communities of color in front-line and essential jobs, one contributing factor may be knowledge gaps around Covid-19 symptoms and transmission. To date, few public health messages around the coronavirus pandemic directly address communities of color and, as a result, these communities may be left behind in the nationwide efforts to contain, respond to, and recover from the pandemic.

The study evaluated the effectiveness of three different video messages about the coronavirus that varied in three ways: the featured physician’s race or ethnicity, the degree to which the physician acknowledged the difficulties faced by communities of color in accessing health services, and messaging around community perceptions of wearing a mask. Physicians from Massachusetts General Hospital (MGH) and Lynn Community Health Center of different racial and ethnic backgrounds were featured in the study’s videos.

“Our motivation to pursue this work stems out of the significant racial and ethnic disparities in access to and take-up of health services that have been widened and intensified by Covid-19,” says Marcella Alsan, professor of public policy at Harvard Kennedy School and co-chair of J-PAL North America’s Health Care Delivery Initiative, part of MIT’s Abdul Latif Jameel Poverty Action Lab.

The findings underscore the importance of physician-delivered messages in increasing knowledge about the virus’ prevention, symptoms, and transmission. They also echo findings of other research demonstrating that race concordance between a patient and physician can increase take-up of health services for Black individuals. For Latinx individuals included in the study, watching a video with a ethnically concordant (Latinx) physician did not impact information-seeking behavior.

As the authors of the study note, increasing knowledge about Covid-19 is only a first step to decreasing transmission and mitigating the negative impacts of the virus. However, given the large gaps in knowledge about Covid-19 among Black and Latinx communities, ensuring that critical public health information is effectively reaching those most impacted by the pandemic is of paramount importance. Increasing the racial diversity of our physician workforce is a longer-term strategy to decrease persistent health inequities, both in the age of Covid-19 and after the worst impacts of the pandemic subside.

de MIT News https://ift.tt/37yR7qD

New “metalens” shifts focus without tilting or moving

Polished glass has been at the center of imaging systems for centuries. Their precise curvature enables lenses to focus light and produce sharp images, whether the object in view is a single cell, the page of a book, or a far-off galaxy.

Changing focus to see clearly at all these scales typically requires physically moving a lens, by tilting, sliding, or otherwise shifting the lens, usually with the help of mechanical parts that add to the bulk of microscopes and telescopes.

Now MIT engineers have fabricated a tunable “metalens” that can focus on objects at multiple depths, without changes to its physical position or shape. The lens is made not of solid glass but of a transparent “phase-changing” material that, after heating, can rearrange its atomic structure and thereby change the way the material interacts with light.

The researchers etched the material’s surface with tiny, precisely patterned structures that work together as a “metasurface” to refract or reflect light in unique ways. As the material’s property changes, the optical function of the metasurface varies accordingly. In this case, when the material is at room temperature, the metasurface focuses light to generate a sharp image of an object at a certain distance away. After the material is heated, its atomic structure changes, and in response, the metasurface redirects light to focus on a more distant object.

In this way, the new active “metalens” can tune its focus without the need for bulky mechanical elements. The novel design, which currently images within the infrared band, may enable more nimble optical devices, such as miniature heat scopes for drones, ultracompact thermal cameras for cellphones, and low-profile night-vision goggles.

“Our result shows that our ultrathin tunable lens, without moving parts, can achieve aberration-free imaging of overlapping objects positioned at different depths, rivaling traditional, bulky optical systems,” says Tian Gu, a research scientist in MIT’s Materials Research Laboratory.

Gu and his colleagues have published their results today in the journal Nature Communications. His co-authors include Juejun Hu, Mikhail Shalaginov, Yifei Zhang, Fan Yang, Peter Su, Carlos Rios, Qingyang Du, and Anuradha Agarwal at MIT; Vladimir Liberman, Jeffrey Chou, and Christopher Roberts of MIT Lincoln Laboratory; and collaborators at the University of Massachusetts at Lowell, the University of Central Florida, and Lockheed Martin Corporation.

A material tweak

The new lens is made of a phase-changing material that the team fabricated by tweaking a material commonly used in rewritable CDs and DVDs. Called GST, it comprises germanium, antimony, and tellurium, and its internal structure changes when heated with laser pulses. This allows the material to switch between transparent and opaque states — the mechanism that enables data stored in CDs to be written, wiped away, and rewritten.

Earlier this year, the researchers reported adding another element, selenium, to GST to make a new phase-changing material: GSST. When they heated the new material, its atomic structure shifted from an amorphous, random tangle of atoms to a more ordered, crystalline structure. This phase shift also changed the way infrared light traveled through the material, affecting refracting power but with minimal impact on  transparency.

The team wondered whether GSST’s switching ability could be tailored to direct and focus light at specific points depending on its phase. The material then could serve as an active lens, without the need for mechanical parts to shift its focus.

“In general when one makes an optical device, it’s very challenging to tune its characteristics postfabrication,” Shalaginov says. “That’s why having this kind of platform is like a holy grail for optical engineers, that allows [the metalens] to switch focus efficiently and over a large range.”

In the hot seat

In conventional lenses, glass is precisely curved so that incoming light beam refracts off the lens at various angles, converging at a point a certain distance away, known as the lens’ focal length. The lenses can then produce a sharp image of any objects at that particular distance. To image objects at a different depth, the lens must physically be moved.

Rather than relying on a material’s fixed curvature to direct light, the researchers looked to modify GSST-based metalens in a way that the focal length changes with the material’s phase.

In their new study, they fabricated a 1-micron-thick layer of GSST and created a “metasurface” by etching the GSST layer into microscopic structures of various shapes that refract light in different ways.

“It’s a sophisticated process to build the metasurface that switches between different functionalities, and requires judicious engineering of what kind of shapes and patterns to use,” Gu says. “By knowing how the material will behave, we can design a specific pattern which will focus at one point in the amorphous state, and change to another point in the crystalline phase.”

They tested the new metalens by placing it on a stage and illuminating it with a laser beam tuned to the infrared band of light. At certain distances in front of the lens, they placed transparent objects composed of double-sided patterns of horizontal and vertical bars, known as resolution charts, that are typically used to test optical systems.

The lens, in its initial, amorphous state, produced a sharp image of the first pattern. The team then heated the lens to transform the material to a crystalline phase. After the transition, and with the heating source removed, the lens produced an equally sharp image, this time of the second, farther set of bars.

“We demonstrate imaging at two different depths, without any mechanical movement,” Shalaginov says.

The experiments show that a metalens can actively change focus without any mechanical motions. The researchers say that a metalens could be potentially fabricated with integrated microheaters to quickly heat the material with short millisecond pulses. By varying the heating conditions, they can also tune to other material’s intermediate states, enabling continuous focal tuning.

“It’s like cooking a steak — one starts from a raw steak, and can go up to well done, or could do medium rare, and anything else in between,” Shalaginov says. “In the future this unique platform will allow us to arbitrarily control the focal length of the metalens.”

de MIT News https://ift.tt/3uoKD7U

Researchers develop speedier network analysis for a range of computer hardware

Graphs — data structures that show the relationship among objects — are highly versatile. It’s easy to imagine a graph depicting a social media network’s web of connections. But graphs are also used in programs as diverse as content recommendation (what to watch next on Netflix?) and navigation (what’s the quickest route to the beach?). As Ajay Brahmakshatriya summarizes: “graphs are basically everywhere.”

Brahmakshatriya has developed software to more efficiently run graph applications on a wider range of computer hardware. The software extends GraphIt, a state-of-the-art graph programming language, to run on graphics processing units (GPUs), hardware that processes many data streams in parallel. The advance could accelerate graph analysis, especially for applications that benefit from a GPU’s parallelism, such as recommendation algorithms.

Brahmakshatriya, a PhD student in MIT’s Department of Electrical Engineering and Computer Science and the Computer Science and Artificial Intelligence Laboratory, will present the work at this month’s International Symposium on Code Generation and Optimization. Co-authors include Brahmakshatriya’s advisor, Professor Saman Amarasinghe, as well as Douglas T. Ross Career Development Assistant Professor of Software Technology Julian Shun, postdoc Changwan Hong, recent MIT PhD student Yunming Zhang PhD ’20 (now with Google), and Adobe Research’s Shoaib Kamil.

When programmers write code, they don’t talk directly to the computer hardware. The hardware itself operates in binary — 1s and 0s — while the coder writes in a structured, “high-level” language made up of words and symbols. Translating that high-level language into hardware-readable binary requires programs called compilers. “A compiler converts the code to a format that can run on the hardware,” says Brahmakshatriya. One such compiler, specially designed for graph analysis, is GraphIt.

The researchers developed GraphIt in 2018 to optimize the performance of graph-based algorithms regardless of the size and shape of the graph. GraphIt allows the user not only to input an algorithm, but also to schedule how that algorithm runs on the hardware. “The user can provide different options for the scheduling, until they figure out what works best for them,” says Brahmakshatriya. “GraphIt generates very specialized code tailored for each application to run as efficiently as possible.”

A number of startups and established tech firms alike have adopted GraphIt to aid their development of graph applications. But Brahmakshatriya says the first iteration of GraphIt had a shortcoming: It only runs on central processing units or CPUs, the type of processor in a typical laptop.

“Some algorithms are massively parallel,” says Brahmakshatriya, “meaning they can better utilize hardware like a GPU that has 10,000 cores for execution.” He notes that some types of graph analysis, including recommendation algorithms, require a high degree of parallelism. So Brahmakshatriya extended GraphIt to enable graph analysis to flourish on GPUs.

Brahmakshatriya’s team preserved the way GraphIt users input algorithms, but adapted the scheduling component for a wider array of hardware. “Our main design decision in extending GraphIt to GPUs was to keep the algorithm representation exactly the same,” says Brahmakshatriya. “Instead, we added a new scheduling language. So, the user can keep the same algorithms that they had before written before [for CPUs], and just change the scheduling input to get the GPU code.”

This new, optimized scheduling for GPUs gives a boost to graph algorithms that require high parallelism — including recommendation algorithms or internet search functions that sift through millions of websites simultaneously. To confirm the efficacy of GraphIt’s new extension, the team ran 90 experiments pitting GraphIt’s runtime against other state-of-the-art graph compilers on GPUs. The experiments included a range of algorithms and graph types, from road networks to social networks. GraphIt ran fastest in 65 of the 90 cases and was close behind the leading algorithm in the rest of the trials, demonstrating both its speed and versatility.

GraphIt “advances the field by attaining performance and productivity simultaneously,” says Adrian Sampson, a computer scientist at Cornell University who was not involved with the research. “Traditional ways of doing graph analysis have one or the other: Either you can write a simple algorithm with mediocre performance, or you can hire an expert to write an extremely fast implementation — but that kind of performance is rarely accessible to mere mortals. The GraphIt extension is the key to letting ordinary people write high-level, abstract algorithms and nonetheless getting expert-level performance out of GPUs.”

Sampson adds the advance could be particularly useful in rapidly changing fields: “An exciting domain like that is genomics, where algorithms are evolving so quickly that high-performance expert implementations can’t keep up with the rate of change. I’m excited for bioinformatics practitioners to get their hands on GraphIt to expand the kinds of genomic analyses they’re capable of.”

Brahmakshatriya says the new GraphIt extension provides a meaningful advance in graph analysis, enabling users to go between CPUs and GPUs with state-of-the-art performance with ease. “The field these days is tooth-and-nail competition. There are new frameworks coming out every day,” He says. But he emphasizes that the payoff for even slight optimization is worth it. “Companies are spending millions of dollars each day to run graph algorithms. Even if you make it run just 5 percent faster, you’re saving many thousands of dollars.”

This research was funded, in part, by the National Science Foundation, U.S. Department of Energy, the Applications Driving Architectures Center, and the Defense Advanced Research Projects Agency.

de MIT News https://ift.tt/2NKNN4Q