Skip to Content

CSE News

  • NSF CAREER Award to Fund Multiplane Data Center Network Research and Education

    University of California San Diego computer scientist George Porter is among the latest recipients of Faculty Early Career Development (CAREER) awards from the National Science Foundation (NSF). The funding agency has earmarked nearly $700,000 for a four-year project on the topic, "A Scalable Multiplane Data Center Network". Funding kicked off on May 15, 2016.

    Porter (at right) is the Co-Director of UC San Diego's Center for Networked Systems (CNS), in addition to being a professor in the Computer Science and Engineering (CSE) department of the university’s Jacobs School of Engineering.

    NSF is committed to promoting the role of teacher-scholars, and the CAREER program is the funding agency’s most prestigious category of awards in support of junior faculty who exemplify excellence in both teaching and research.

    As Porter outlined in his proposal, "Supporting the ever-increasing data rates required between literally hundreds of thousands of servers is an extremely challenging problem... [and] solving this problem is critical to building and deploying compute clusters capable of meeting the exponentially growing requirements of users and developers building new applications."

    He noted that industrial data center operators such as Google and Microsoft are relying on scale-out designs that are prohibitive in cost and "power-hungry". As a result, said Porter, "inefficiencies due to network bottlenecks get magnified across hundreds of thousands of servers, resulting in huge costs." One way around this logjam is the design of hybrid networks based on reconfigurable physical optical circuits from one part of the network to another, because such circuits would not require expensive conversion of optical signals into electric signals and then back to optical. Nor do they require transceivers or intermediate packet switches.

    "However, due to technological and physical limitations, for next-generation data center bandwidths, we cannot simply 'speed up' existing hybrid designs," explained Porter. "We eschew the idea of designing a single network fabric that tries to both support next-generation bandwidths and scale to thousands of endpoints," explained Porter. "Instead, we propose a composite network fabric built from multiple, entirely physically independent sub-networks... that can each sale to thousands of nodes, yet are, by themselves, not able to meet the end-to-end bandwidth demands of the data center." To meet those bandwidth needs, he added, traffic can be rapidly switched between the sub-networks.

    Porter (pictured at left with fellow CNS co-director Stefan Savage) argues that using multiplane network nodes as the building blocks for end-to-end data center network topologies -- instead of using conventional switching -- can greatly reduce the number of switches, lower cost and power consumption, resulting in faster link rates. "A given set of sub-networks do not necessarily offer a circuit between each endpoint, meaning that some data must transit across a number of intermediate points until it reaches the ultimate destination," noted Porter. "This indirection approach, coupled with physically separate sub-networks, is a unique aspect of our proposed systems research."

    Education is a critical component of CAREER awards, and Porter says that he will translate the research of his project into new and existing courses based on hands-on projects and system-building experiences, at both the undergraduate and graduate levels. For undergrads, he is planning a networking 'maker' course, and he'll continue to mentor teams of students in UC San Diego's Early Research Scholars Program (ESRP), which emphasizes underrepresented minority students and women getting involved in the research process. He also plans to create a series of videos for UCSD-TV to "provide the public and pre-college students with context for the challenges presented by powering the Internet and cloud computing, and highlighting the importance of the research in this area."

  • CSE Faculty Welcome White House National Microbiome Initiative

    The White House Office of Science and Technology Policy (OSTP), in collaboration with public- and private-sector stakeholders, rolled out a National Microbiome Initiative (NMI) on May 13. Its goal: to foster the integrated study of microbiomes — the communities of microorganisms that live on and in people, plants, oceans and the atmosphere — and to develop tools to protect and restore healthy microbiome function.

    On hand for the announcement was Rob Knight, a professor of Pediatrics and Computer Science and Engineering, who has been closely involved in the effort to create a national microbiome strategy since before he arrived at UC San Diego in early 2015. In announcing the NMI, the White House also noted that UC San Diego will invest $12 million in its newly-created Center for Microbiome Innovation (CMI) to “enable technology developers to connect with end users” of microbiome-related technology. CMI is directed by Knight and involves CSE faculty including Pavel Pevzner, Vineet Bafna, Nuno Bandeira, and Larry Smarr among its 150 faculty researchers. 

    “Microbes pervade all kinds of processes — from our bodies to our planet to industrial fermentation and drug synthesis,” said Knight (pictured at left with Smarr), who also directs the UC San Diego-based American Gut crowdsourcing project, which will expand its partnerships to improve public and academic understanding of the microbiome, as noted in the May 13 announcement. “Working closely with other researchers in the White House’s National Microbiome Initiative will help us unravel the fundamental science so we can understand how microbes do all these things, and help us improve the speed and accuracy in which we can ‘read out’ microbes.”

    Smarr, who also directs the California Institute for Telecommunications and Information Technology (Calit2), recalls that the genesis of the institute’s research on microbiomes dates to January 2006 and a $24.5 million effort funded by the Gordon and Betty Moore Foundation. Called the Community Cyberinfrastructure for Advanced Marine Microbial Ecology Research and Analysis (CAMERA), the effort was led by Smarr and over seven years it built a computational resource at UC San Diego to help scientists understand how microbes function in their natural ecosystems, a field called metagenomics.

    (Smarr and Knight, pictured at left, appeared recently in a conversation on "Decoding the Microbiome", part of CSE's Computing Primetime series on UCSD-TV. Click here to watch the program online.)

    Even before the seven-year CAMERA project ran its course, Smarr made a major commitment to understanding the practical – and personal – applications of microbiome science as part of his Future Health project. The effort began with Smarr’s decision to study his own microbiome and the role it plays in controlling or exacerbating his Crohn’s disease. Smarr also collaborates with Rob Knight, including on a multidisciplinary CMI project to sequence the gut microbiomes of people with inflammatory bowel disease (IBD), such as Crohn’s disease, affecting Smarr and millions of Americans.

    According to Smarr, UC San Diego researchers "are providing new platforms for the study of microbiomes, notably through our advances in scientific visualization and ultra-broadband networking through the Pacific Research Platform." 

    Enabling the effort to build platforms, Knight and his team developed Quantitative Insights into Microbial Ecology, or QIIME (pronounced chime), a microbiome analysis platform distributed through Illumina’s BaseSpace app store. Both the American Gut Project and the NIH-led Human Microbiome Project now rely heavily on the open-source QIIME software, as does Calit2’s Smarr. “We look forward to using QIIME on BaseSpace for our upcoming deep dive into the differences in the human gut microbiome in healthy people compared to people with inflammatory bowel disease,” Smarr said when the partnership with Illumina and BaseSpace was announced.  

    Visit the website of UCSD's Center for Microbiome Innovation.

  • Milestone Reached by CSE Startup as Newspaper Features Mobile App

    The headline in the Union-Tribune newspaper says it all: "App makes it easier for geeks to socialize at events."  The headline sums up the latest startup of CSE Prof. Yuanyuan (YY) Zhou and former students, as the company reports a key milestone for the maker of an app to foster event engagement and networking.

    Zhou is the co-founder and CEO of Whova, a company that created an app for smartphones that can transform the way that researchers and other event-goers network with fellow attendees. In its first full year on the market, the Whova app has now been used at over 3,000 events in 80 countries. That means more than 350,000 conference attendees could use the app to network more effectively at trade meetings and academic conferences and events (including at CSE's 25th anniversary celebration earlier this year).

    In the article, updated on May 9, professor Zhou told reporter Mike Freeman that Whova is "using geeky ways to help geeks like us to know how to socialize." The newspaper also notes that Whova has raised $6 million in venture-capital firms and angel investors -- enough funding to support an operation that has grown to roughly 20 full-time employees, including CSE alumni Jiaqi Zhang (Ph.D. '14) and Zhuoer (Joel) Wang (Ph.D. '14). 

    Like other mobile apps serving the event market, Whova conveniently incorporates a conference agenda and maps of each conference or exhibition area, but Zhou is particularly proud of the advanced features, including analytics technology, that Whova pioneered. The service gets the names and email addresses of registered attendees from the conference organizer, then uses its data analytics technology to generate a networking profile for each person based on all existing information about him or her on the Internet. Whova "provides event attendees insights about each other so that they can plan in advance whom they want to meet at the event," Zhou told the newspaper, which notes that the app also allows attendees to chat and send instant messages, or use the camera function to gather and index the business cards of other attendees. 

    Freeman also spoke with Whova co-founder Weiwei Xiong (pictured second from left after Zhou), a former visiting student in Zhou's group in CSE who also worked for her at the University of Illinois at Urbana-Champaign (UIUC), where Zhou taught before joining the UC San Diego faculty in 2009. "We focused on event engagement and networking and created a lot of unique features that competitors never had," explained Xiong, "and it is hard for them to catch up." Xiong worked as a student in Zhou's San Diego group for three years starting in 2009 even as he was finishing his Ph.D. at UIUC in 2012. Zhou and Xiong began developing Whova the following year, but they didn't begin charging for the service until late 2014. Another co-founder of Whova, Soyeon Park (far right), was a postdoctoral researcher at UIUC who followed Zhou to UC San Diego in 2009 to work in her group prior to joining Whova full-time, as did another co-founder, Tianwei (Tim) Sheng (second from right), who was a postdoc in Zhou's CSE group until joining Whova in late 2013.

    Read the full article in the San Diego Union-Tribune.

  • Honorable Mention to Bioinformatics Professor's Doctoral Dissertation

    ECE Prof. Siavash Mirarab (left) is also a member of the joint Bioinformatics program with CSE and other UC San Diego departments (along with CSE faculty, including Vineet Bafna, Nuno Bandeira and Pavel Pevzner), so it's worth noting that he was a runner-up for the ACM Doctoral Dissertation Award. The award and two Honorable Mention awards were just announced and will be handed out at a banquet June 11 in San Francisco.  Prof. Mirarab received an Honorable Mention for his dissertation on "Novel scalable approaches for multiple sequence alignment and phylogenomic reconstruction." He was nominated for the top award by his alma mater, the University of Texas at Austin, where he finished his doctorate before joining the ECE faculty in September 2015.

    Mirarab’s dissertation addressed the growing need to analyze large-scale biological sequence data efficiently and accurately. To address this challenge, he introduces several methods: PASTA, a scalable and accurate algorithm that can align data sets up to one million sequences; statistical binning, a novel technique for reducing noise in estimation of evolutionary trees for individual parts of the genome; and ASTRAL, a new summary method that can run on 1,000 species in one day and has outstanding accuracy. These methods were essential in analyzing very large genomic datasets of birds and plants.

  • UC San Diego Ranked #35 in Prestigious Global Ranking

    According to the world's largest, invitation-only opinion survey of senior academics, UC San Diego is the #35 best university around the globe. Just out, the UK-based Times Higher Education Supplement released its World Reputation Rankings in early May. The campus's reputation climbed six spots compared to its ranking released in 2015.

    “We are pleased UC San Diego’s reputation continues to grow as a world-class university that provides an outstanding education and pushes the frontiers of knowledge to improve our local and global communities,” said Chancellor Pradeep K. Khosla. “UC San Diego’s ranking as a global leader is a reflection of our dynamic students, faculty and staff, who foster a culture of collaboration, spark discoveries and transform lives.” According to the Times Higher Education announcement, the 2016 World Reputation Rankings "employ the world’s largest invitation-only academic opinion survey to provide the definitive list of the top 100 most powerful global university brands, based on the judgment of senior, published academics––the people, the organization believes, who are best placed to know the most about excellence in universities.”

    UC San Diego has dominated academic world rankings in recent months. The campus was ranked the world’s 14th best university for the third consecutive year by the 2015 Academic Ranking of World Universities (ARWU). In a separate ranking, the Center for World University Rankings put the campus at #21, and U.S. News & World Report gave UC San Diego the #19 spot in its latest global ranking of universities.

    The Times Higher Education survey did not attempt to uncover the individual departments or disciplines with the best reputations. But in its latest ranking, U.S. News put computer science at UC San Diego at #11 among the best global university programs, and the last ARWU survey ranked the CSE program at #14 worldwide.

  • CSE Neuroengineer Tackles Consciousness, Neuromorphic Engineering and Machine Learning

    Computer scientists are not often invited to present their research at The Science of Consciousness annual conference, but University of California San Diego development engineer Stephen Deiss did just that. He spoke to the meeting in Tucson, AZ, in late April on the subject of "Romancing the Oxymoron: The 'Hardware Problem' of Machine Consciousness."

    "I presented my view that our preconceptions – about causality and mechanisms – bias us against accepting the possibility that machines can be conscious," said Deiss (at left), who earned his M.S. in computer science at Purdue University. "It leaves us believing that there is something spooky and unnatural about our own awareness, but I argue that consciousness is fundamental and scale-free in nature."

    Supported by CSE since 2013, Deiss first worked with the Non-Volatile Systems Laboratory of CSE Prof. Steven Swanson, but now splits his time between the Integrated Systems Neuroengineering Lab of Bioengineering professor Gert Cauwenberghs, and the new Pattern Recognition Laboratory of the Qualcomm Institute (both with CSE support). "We are focused on neurally-inspired or otherwise non-von Neumann computing paradigms," said Deiss, referring to new architectures that, unlike most of today’s computers, are not based on executing instructions sequentially. The scientist is looking particularly for neural applications to neuromorphic engineering and machine learning.

    In his presentation, Deiss noted that the "presumption that engineers cannot give [qualitative sensory and cognitive] sensations to machines is a reason many refuse to entertain the possibility of machine consciousness." He added that theories of consciousness have grown more sophisticated and quantitative. "I have maintained for over a decade that the crux of the problem is the assumption that there are laws operating on nature from a higher mathematical or divine realm," explained Deiss. "If one is able to abandon this view for a radically secular view of nature, the question becomes how natural systems do what they do from intrinsic principles and constraints rather than as externally directed."

    The computer scientist argues that neuromorphic engineering, deep learning and other methods, as well as models based on free-energy theory and Bayesian inference, inevitably lead to the engineering of machines that can do more than we can. "I argue that all manner of systems, from atoms to brains and beyond, are conscious, with highly variable perceptual skills and a spectrum of self-reference," said Deiss. "Natural systems sense, detect and interpret when they interact and thereby assign meaning, but that is the same process that goes on in our predictive brains. I also argue that causal modeling is a heuristic that works, but it leads us astray in thinking about consciousness."

    Deiss's ruminations on machine (and human) consciousness have deep roots, starting with dual undergraduate majors in psychology and philosophy at the University of Michigan. His first job out of graduate school involved applications of artificial intelligence, cognitive and computer science to computer-aided instruction. He went on to do hardware and software engineering in satellite communications, then high-energy physics, before beginning a 16-year career designing platforms for artificial neural networking. Ultimately, he gravitated to neuromorphic engineering (he calls himself a neuroengineer), and Deiss admits that his viewpoint on consciousness is a radical one: "A paradigm shift regarding consciousness is ultimately required -- and it's on the way."

    Read the abstract for Stephen Deiss's presentation at The Science of Consciousness 2016.

  • Interested in UX Design? CSE's Klemmer Has Two of the Top Three MOOCs in New List

    According to a feature article in the British online publication Digital Arts, an online course developed and taught by CSE (and Cognitive Science) Prof. Scott Klemmer is one of the best in the world for learning user experience (UX) design.

    The survey noted that "UX design is all about improving the user experience through creating easy-to-use products that are a pleasure to use, and customizing a product or service to take into account exactly how the user interacts with" that product. The article breaks down available UX design courses between massive open online courses (MOOCs), and at-your-own-pace software packages from the likes of (owned by LinkedIn), Udemy, Treehouse, Pluralsight and Skillshare (where you can choose to be either a teacher or a student).

    The Digital Arts review focuses on Coursera as the primary MOOC platform for both paid and free online courses. At the top of the list was the popular Interaction Design Specialization, a series of seven courses taught by Klemmer and a capstone design project that, taken together, promise to help students "learn how to design great user experiences." "These courses are a brilliant overview on a wide range of UX aspects, and include quizzes and assignments to keep you engaged, as well as great teachers such as Scott Klemmer," writes Mimi Launder in "The 11 Best Paid and Free UX Design Courses". "These are resource-rich, tough courses that -- if you complete them -- you will gain a lot from." The article also singles out Klemmer's Human-Computer Interaction course on Coursera as one of the best free online UX courses. Klemmer's former employer, Stanford, still offers his original MOOC on Coursera free of charge. According to the reviewer, "it may not look quite as sleek as the paid course [from UC San Diego], but, well, it is free. And it offers brilliant video lectures, though none of the assignments or quizzes, unlike the paid version" available on Coursera from UC San Diego.

  • Graduate Student Community Awards to CSE Support Staff and Community Leader

    Jessica Gross is the TA and Graduate Admissions Coordinator in the CSE Student Affairs, and she is one of two winners for the department at the Graduate Student Association (GSA) Community Awards. Meanwhile, Ph.D. candidate Dustin Richmond -- who expects to graduate this June -- was also singled out for an award: the GSA Community Award for Outstanding Community Leader.

    Gross (at right) won the Community Award for Graduate Student Support Staff. She oversees the TA and tutor programs in CSE as well as graduate admissions and recruitment for the Ph.D. program. The Support Staff Award honors UC San Diego staff who go "above and beyond" their job requirements in assisting graduate students. According to CSE Prof. Sorin Lerner, Gross works "tirelessly on TA-assignments, Ph.D. admission and Ph.D. Visit Day, this year organizing the visit of over 40 prospective Ph.D. students, including assisting them with travel and accommodations," said CSE Prof. Sorin Lerner. She was also involved with the renovation of the graduate student lounge (Chez Bob) and regularly helps with Social Hour. Gross also attends and contributes to many meetings of the Graduate Community Council, a department group dedicated to improving graduate community in the CSE department. Gross is a UC San Diego alumna, having received her B.A. in Linguistics in 2012. She began working as a graduate program assistant in CSE while still an undergraduates, and began working as an intake advisor in Student Affairs, also in the CSE department.

    In awarding the Outstanding Community Leader honor to Dustin Richmond (at left), the GSA cited his role for several years in a row in leading graduate-student volunteers for Visit Day. In 2015-16, he also led the student-faculty candidate meetings and evaluations, which resulted in very thoughtful and well-organized comments. Richmond was the key instigator for the re-do of the Chez Bob graduate lounge and he played a huge leadership role in making those renovations happen. As the recipient of an NSF Graduate Research Fellowship in 2012, Richmond was also cited for having taken upon himself the responsibility of organizing several workshops to help fellow CSE graduate students learn how to apply successfully for NSF fellowships. He has been pursuing his Ph.D. in CSE since 2012, prior to which he did simultaneous undergraduate degrees in Computer Engineering and Electrical Engineering at the University of Washington. Under his advisor, CSE Prof. Ryan Kastner, Richmond is designing an ultra-high-speed image processing pipeline for active 3D scanners using compressive-sensing techniques. He is also the lead designer for an FPGA system to decompress and process 20,000-plus images per second.

  • CSE Sophomores Take VR Skills to Santa Barbara Hackathon

    UC San Diego computer science sophomores Connor Smith, Kristin Agcaoili and Anish Kannan were on hand April 22-24 when UC Santa Barbara hosted the second annual Santa Barbara Hackathon. The campus group SB Hacks hosted the 36-hour marathon coding competition, which was open to college students from around California, including community college students.

    CSE was represented by (pictured l-r) Smith, Agcaoili and Kannan, all members of the Virtual Reality Club at UC San Diego. The team created Chemistry Lab VR, an educational virtual-reality experience that teaches students lab procedure and safety. It could be used in chemistry classrooms to simulate real-life lab procedures to reduce the risks of working with potentially harmful chemicals. This was not the first time that the VR Club team developed a program to improve scientific instruction: at the HackingEDU hackathon in October 2015, Smith, Agcaoili and Kannan successfully coded a Cell VR program to teach cell biology and interact with a virtual human cell. They finished in third place at HackingEDU. The same team of three students also competed in November 2015 at HackSC in Los Angeles, where their application Diver -- to spread awareness about ocean pollution -- was awarded the Best VR/Game Hack of the USC-organized hackathon.

    As Smith told the UC Santa Barbara campus newspaper in an interview, the team traveled from San Diego to make a statement. "There are prizes, but we don't really do it competitively," he said. "We're just trying to do something that has impact beyond this space." The team arrived at UCSB with an HTC Vive system that Smith described as a "new, room-scale virtual reality, where you can actually walk, move and duck, all within the given boundaries."  Vive users are outfitted with a headset, headphones and two handheld controllers, and they interact with a programmable interface that allows "free movement between two lighthouse motion trackers that define a 16-square-foot space." (In 2015, Smith was an HTC Brand Ambassador, which paid him to demonstrate the VR system to fellow students.)

    Smith told the campus newspaper that he couldn't have asked for more support from hackathon organizers. "It's just a really supportive environment," he noted. "There are mentors, there's free food, great energy and just a lot of people working on something really cool. These events output so many cool projects as well, things that maybe people wouldn't have time to do otherwise.

  • Andrew Kahng's Best Paper at CDNLive Silicon Valley and Megatrends at Cadence Distinguished Speakers Series

    For the very first time, CDNLive Silicon Valley ran an academic track during its 2016 meeting, with a UC San Diego professor taking top honors. CDNLive is the Cadence User Conference, which attracts experts as well as users and developers of Cadence technologies and techniques for designing advanced silicon, systems, and systems-on-chip. The April 5-6 event took place in Santa Clara, CA, and drew a standing-room-only audience. According to a conference rapporteur, CSE and ECE Prof. Andrew Kahng "stole the show" with his talk on bridging the gap between academic research and commercial Electronic Design Automation (EDA)."  To noone's surprise, the presentation, "Toward New Synergies Between Academic Research and Commercial EDA",  received the Best Paper award for the Academic track, which included presentations by faculty from six other universities: USC, Arizona State, Georgia Tech, Purdue, Texas A&M, and the University of Calgary.

    On the day after he spoke at CDNLive, Kahng (at right) presented as part of Cadence's internal Distinguished Speakers Series. His talk focused on "PPAC Scaling at 7nm and Below."  PPAC stands for power-performance-area-cost. The industry has talked about PPA for a long time, with the A for area also being a surrogate for cost. But with different process choices, multiple patterning vs. EUV someday and other options, area alone is not the only parameter that feeds into cost. According to a review of Kahng's talk by Cadence's Paul McLellan, the UC San Diego professor sees two megatrends that are driving "all the issues" in this area. The first is the "race to the end of the roadmap", i.e., taking Moore's Law as far as possible on existing or foreseeable knowledge, which is expected to result in volume production of 7nm technology in 2018. The second megatrend involves keeping power under control by adopting more extreme approaches after having "done a lot of the easy stuff in previous process generations" to enable low power for markets ranging from mobile to big data to cloud. Kahng's final call to arms in his Cadence talk was for a massive "moonshot" to predict tool outcomes, find the sweet spot for different tools and flows, and thus design in specific tool and flow knobs to the overall methodology. The end-result: a fully predictive, one-pass flow with optimal tool usage. With modern massively parallel, big-data architectures, argued Kahng, who is a former member of the Cadence Technology Advisory Board, it is not unreasonable to use tens of thousands of machines if it could "get us to the moon" of a non-iterative flow.