08.04.17

 Friday, August 4, 2017
 

New Computer Engineering Professor Crosses Fields of Computer Architecture and System Software

She won't take up her new assistant professorship at the University of California San Diego until January, but Jishen Zhao is the latest hire into the university's Computer Science and Engineering (CSE) department. Currently an assistant professor of computer engineering at the University of California, Santa Cruz, Zhao intends to expand on her prior work in computer architecture and software systems, with a particular emphasis on memory and storage systems, domain-specific acceleration, high-performance computing and energy efficiency. Other research interests include electronic design automation and VLSI design for three-dimensional integrated circuits and nonvolatile memories.

"The fields of computer architecture and systems have seen tremendous achievements in preserving continuous performance growth of computers," said Zhao. "Now two major challenges remain: the complexity and richness of application demands continue to overwhelm hardware capabilities, and new physical devices and circuit technologies are developed with properties that disrupt previous design assumptions - thus making current computer system designs suboptimal."

Prior to joining the UC Santa Cruz Baskin School of Engineering faculty in 2015, Zhao was a research scientist at HP Labs, where she worked in both the Systems Research Lab and HP's Intelligent Infrastructure Lab. She joined the company after completing her Ph.D. in Computer Science and Engineering at Pennsylvania State University in 2014. She received her undergraduate and M.S. degrees in Electrical Engineering from China's Zhejiang University in 2003 and 2008, respectively.

Professor Zhao's honors include an NSF CAREER Award (2017), which came with nearly $500,000 over five years to pursue her research on enhancing data center storage systems with persistent memory. Persistent memory was also the focus of her seminar at UC San Diego in April, when she laid out a hardware-driven approach to "unlocking the full potential of persistent memory."  Persistent memory was also the underlying topic that Zhao addressed in a paper to the Design Automation Conference (DAC 2017) earlier this summer in Austin, TX.  Co-authored by Zhao with scientists from National Taiwan University and Academia Sinica, the joint paper explored a new persistent memory-accelerating approach.*

Zhao's other honors include an honorable mention for the best-paper award at the International Symposium on Microarchitecture (MICRO 2013). The paper presents the first hardware-based persistent memory design that outperforms all previous software-based solutions.

Her latest publications include two papers accepted for presentation at the 2017 International Conference on Computer Aided Design (ICCAD 2017) in Irvine, CA. Both relate to magnetic random access memory (MRAM): "Approximate Image Storage with Multi-level Cell STT-MRAM Main Memory" and "PRESCOTT: Preset-based Cross-Point Architecture for Spin-Orbit-Torque Magnetic Random Access Memory."

Apart from her previous experience at HP Labs, Professor Zhao brings to UC San Diego a strong set of research relationships with industry, notably companies such as AMD, Intel, Micron, and Western Digital. "My collaborations exposed me to the technical challenges at the cutting edge of future computer systems but also promoted my research impact," explained Zhao. Her work on GPU system design using 3D+2.5D integration technology followed her collaboration with AMD, one of the top GPU vendors. Added Zhao: "Shortly after my work was published, both AMD and NVIDIA began to adopt such technology in their high-bandwidth, memory-based GPU products. My research also contributed nine patents to industry's intellectual property portfolios."

While her work addresses both technologies and applications roughly equally, Zhao admits that modern applications such as web search are usually executed on tens of thousands of interconnected processors. "This calls for a much broader research paradigm beyond stand-alone computers," she noted, "so I plan to explore digital ecosystems that are composed of interconnected sensors, mobile and cloud servers. I would like to research the scalability of my standalone computer designs as well as the problems posed by the heterogeneity of digital ecosystems."

In CSE, Jishen Zhao looks forward to collaborating with faculty in the Non-Volatile Systems Lab, including Steven Swanson. Continuing her dual focus on applications and technologies, she sees research directions driven by advancements on each count. With new technology (in the form of persistent memory) expected to take its place as a new tier in a data center's memory/storage stack because of their promising density and cost, Zhao plans to study next-generation designs that aim to use persistent memory in the cloud. "It's promising," she noted. "But this technical transition fundamentally changes current memory and storage system design assumptions... and introduces critical design challenges that demand a full-scale rethinking of data center memory and storage systems."

Read the complete news release on the CSE website.

____________________________

*C.-H. Lai, J. Zhao and C.-L. Yang, "Leave the Cache Hierarchy Operation As It Is: A New Persistent Memory Accelerating Approach", Design Automation Conference (DAC 2017), June 2017, Austin, TX. 

 

En Route to UC San Diego, Chinese Undergraduate Wins Award at SIGGRAPH 2017 - in Absentia

This fall, an undergraduate alumnus from China's Tsinghua University will be joining the Computer Science and Engineering (CSE) department and the computer vision and graphics group of CSE professor Ravi Ramamoorthi. The student, Tiancheng (Kevin) Sun, will also be active in the Center for Visual Computing that Ramamoorthi directs.

Sun is making a splash even before arriving in San Diego to start the Ph.D. program. He was named winner in the undergraduate category of the ACM Student Research Competition at the 44th SIGGRAPH Conference on Computer Graphics and Interactive Techniques. At the July 30 to August 3 SIGGRAPH conference and expo this week in Los Angeles, Sun presented a poster on "Attribute-Preserving Gamut Mapping of Measured BDRFs." The poster itself was based on a paper published recently at the 2017 EuroGraphics Symposium on Rendering, which took place in June, and in the Computer Graphics Forum  (Vol. 36, Issue 4, July 2017, pp. 47-54).

The Chinese student was first author on the paper and poster, together with Ph.D. student Ana Serrano and professors Diego Gutierrez and Belen Masia (in the Graphics and Imaging Lab of Spain's University of Zaragoza). Sun was a summer intern in the Spanish lab last summer (and later a collaborator). He also did a recent internship in the Visual Computing Center of Saudi Arabia's King Abdullah University of Science and Technology.

While still an undergraduate, Sun was admitted to a special Computer Science undergraduate program established by Tsinghua's Institute for Interdisciplinary Information Sciences (IIIS). The program is known as the "Yao Class" because the institute is led by ACM Turing Award laureate Andrew Yao (who received the top award in Computer Science in 2000).

At SIGGRAPH, Ana Serrano presented the poster because Sun - who completed his B.S. at Tsinghua in July - is still waiting to undergo his background check for a U.S. visa.

Having won the SIGGRAPH student contest, Sun now qualifies to compete in ACM's overall Student Research Competition, alongside other undergraduate student winners at ACM-sponsored conferences that offer student research competitions.

When he enrolls in UC San Diego this fall, Sun will benefit from a Jacobs Fellowship, the most selective university fellowship for incoming Ph.D. students. Meanwhile, he is building a research library of light-field imaging and displays, to include relevant papers and a roadmap to this area of potential future research. (Sun says his primary research interests are computational imaging and display.)

 

CSE Alumnus Demonstrates Rapid but Realistic 3D Avatar Creation

CSE alumnus Iman Sadeghi (M.S., Ph.D. '08, '11), who did his dissertation in computer graphics on controlling the appearance of specular microstructures, is now VP of Engineering for a Santa Monica-based startup. That company, Pinscreen, aims to revolutionize digital communication and online human interaction with user-generated and personalized 3D virtual avatars.

Sadeghi and other Pinscreen engineers staged a high-profile demo for Real-Time Live, an "interactive extravaganza" that was part of SIGGRAPH 2017 in Los Angeles this week. Like other participants, the team from Pinscreen did real-time demos of the company's 3D avatar and tracking software (now in beta release). CSE alumnus Sadeghi led the demo on "Creating Performance-Driven Avatars in Seconds", showing how the software can capture a 2D selfie taken with a webcam and convert it - in a matter of seconds - into a surprisingly realistic 3D avatar "that stops just short of the uncanny valley" (referring to the point at which a robot or avatar is so realistic that a viewer starts to feel uncomfortable with it). The software runs a neural network on a graphics processing unit (GPU). Click here to watch the video on Facebook .

Using the Unity game engine, the Pinscreen team also demonstrated how to insert the rapidly-created 3D avatar into a gaming environment, and showed how the expressive avatar can reflect different light conditions programmed into the game. They also showed 3D avatars draped over participants in real-time videoconferencing.

Sadeghi's colleagues participating in the demo included Koki Nagano, Jaewoo Seo, and Carrie Sun, all from Pinscreen.

 

Professor Behind Online Computer Graphics Course among Finalists for 2017 edX Prize

For the second year in a row, CSE professor Ravi Ramamoorthi -- who holds the  Ronald L. Graham Chair of Computer Science -- is a finalist in a competition for an online education award. The 2017 edX Prize for Exceptional Contributions in Online Teaching and Learning has whittled down all nominations to a list of 10 finalists among the developers of online courses delivered via edX, one of the largest massive open online course platforms in the world.

Professor Ramamoorthi's "Computer Graphics" online course was the first in the field of computer graphics. Originally developed while he was at UC Berkeley, the course is now part of UC San Diego's offerings on the edX platform. In 2016 the course was also a finalist in the annual competition. In the past year, the course has enrolled approximately 20,000 students, bringing to more than 100,000 the number of students who signed up to take the course since it was first offered.

All submissions were reviewed by an evaluation committee made up of faculty, instructional designers, college deans, and online learning directors from edX partner institutions. The evaluation committee selected 10 finalists. The edX University Advisory Board will evaluate the finalists and select the winner of the edX Prize. The award will be announced at the 2017 edX Global Forum hosted by the University of British Columbia in December, when all the finalists will be honored.

The next edition of the self-paced "Computer Graphics" course on edX starts on September 18, 2017.

Computer Graphics on edX

 

CSE and Contextual Robotics Team Returns from First-Ever Appearance at RoboCup@Home Competition

It was the first time a team from UC San Diego's Contextual Robotics Institute competed head-on with top robot builders around the world at the international RoboCup@Home competition in Japan. A team of computer scientists from the institute and CSE failed to make the top three (the only U.S. institution to make the top three was the University of Texas at Austin, which placed #3 in the same category).

The competition ran from July 27 to 31 in Nagoya, and members of the UC San Diego team spent the previous three months refining and testing algorithms to train Toyota's Human Support Robot (HSR) to complete two tasks: picking up and putting away groceries; and helping someone to carry groceries from their car to their home. The students were part of the research groups of CSE professors Laurel Riek and Henrik Christensen. (Christensen is also director of the Contextual Robotics Institute.)

The RoboCup@Home competition tests a robot's ability to perform everyday tasks, help people around the house and establish naturalistic robot-human communication and interaction. The teams were also tested on their ability to control robotic arms to manipulate objects; and the use of a wide range of sensors for voice interaction and image recognition. "This is a great competition because it challenges researchers to think about the practicalities of deploying robots in real-world environments, like people's homes," said professor Riek. "In a very short time, our students have accomplished remarkable work, and we are really proud of them."

The Tritons-SPL team faced off against 10 teams, including UT Austin and Northeastern University. In addition to UT Austin's #3 finish, the top performers in UC San Diego's category were: #1-ranked Hibikino-Musashi@Home SPL, a team fielded by Kyushu Institute of Technology; and #2-ranked eR@sers, a team including members from Tamagawa University and from Japan's National Institute of Informatics, and the National Institute of Information and Communications Technology. 

All teams used Toyota's HSR robot. The robot is designed to assist people with daily tasks. It's compact and light weight, and equipped with an articulated arm and flexible hand as well as a telescoping body. The robot can be programmed and trained to fetch and carry objects, recognize and interact with people, and perform household tasks.

The UC San Diego team was made up of CSE Ph.D. students Angelique Taylor, Darren Chan, Priyam Parashar and Ruffin White. Taylor and Chan are members of the Healthcare Robotics Lab, led by Riek, which focuses on human-robot teaming research in safety-critical environments, such as hospitals, homes and factories. They design learning algorithms for robots which can understand people, and are currently focused on applications in critical care and neurorehabilitation. Riek leads multiple projects supported by the National Science Foundation, Air Force Office of Scientific Research, and industry, which focus on new computational methods for human-robot teaming.

Ph.D. students Parashar and White are members of the Cognitive Robotics Lab, led by Christensen, with a focus on systems integration, human-robot interaction, mapping and robot vision. The ultimate goal is to build robots that empower people in their daily lives.

Read a preview of the robotics competition on the Jacobs School of Engineering website. 
RoboCup@Home Domestic Standard Platform 

 

CSE Fills New and Existing Senior Jobs in External Affairs and Student Affairs

CSE's MSO, Samira Khazai, announced two new midsummer staff hires. 

She welcomed Rebecca (Becky) Hames to be the CSE department's External Relations Manager. Hames fills a new and vital role in the department, developing a vibrant alumni outreach program and facilitating relationships between our students, alumni and researchers as well as industry and community organizations.  In her new role, Hames is collaborating with the many existing units and resources on campus: Central Development, Alumni Relations, Career Center, Corporate Affiliates Program (CAP), the Team Internship Program (TIP), and the Jacobs School Communications unit to connect our students, faculty and researchers with potential educational, recruitment and research project opportunities.  Hames is additionally charged with managing CSE's social media presence and ensuring the timely and accurate dissemination of the department's external communications through newsletters, websites, CSE's Computer Science Channel in conjunction with UCTV, etc.  Becky Hames is open to fresh ideas for alumni-related, industry-related, educational or research projects. She works on the second floor of the CSE building (room 2214) or she can be reached via email

Effective this coming Monday, August 7, Patrick Mallon is joining the department as CSE's new Director of Student Affairs. Noting that the Student Affairs unit has been operating without a director for nearly a year, MSO Khazai welcomed Mallon's arrival, saying that he "brings with him a great deal of experience and enthusiasm for this position." Mallon is moving from the Mechanical and Aerospace Engineering (MAE) department, where he quickly rose through the student affairs staff ranks to serve as their current Director of Student Affairs. He has impressive skills in financial support management, program administration and "a passion for solving intractable student affairs challenges," added Khazai. Mallon's leadership skills and expertise will be especially critical at a time of significant growth in CSE's undergraduate and graduate student populations, and the resulting challenges to managing enrollment, finding adequate instructional resources and scarce classroom space. Mallon's office will be room 1228 in the Student Affairs unit on the first floor of the CSE building. He can be reached at this email address.

 

     UPCOMING EVENTS

MONDAY, AUGUST 7, 2017        

Software Techniques to Enhance Reliability of Emerging Compute and Memory Units

Ph.D. candidate Manish Gupta is set to defend his dissertation on Monday in front of a committee consisting of his co-advisors Rajesh Gupta (no relation) and Dean Tullsen. The panel will also include CSE's Ranjit Jhala and Ryan Kastner, as well as ECE professor Truong Nguyen. Gupta was part of the NSF Variability Expedition, and his dissertation is on "Software Techniques to Enhance Reliability of Emerging Compute and Memory Units." The examination is open to interested members of the UC San Diego community.

Date: Monday, August 7
Time:  10:00 a.m.
Location: Room 3217, CSE Building

Abstract: System reliability is becoming a significant concern as technology continues to shrink. This is because of increasing variation in circuit characteristics seen in nanometer-scaled microsystems. These variations in semiconductor manufacturing manifest as increasing fault rates in devices, rising soft errors, and timing errors caused by accelerated aging in circuits. This dissertation seeks software-based techniques to detect, recover, and prevent such errors in compute and memory components.

Software-based error detection and recovery techniques suffer from high-performance penalty to the overall system. This thesis presents methods that minimize the performance overhead of software-based error mitigation. In particular, this work proposes two techniques for effective software-based error detection in compute units: fingerprinting and cross-lane instruction. Fingerprinting combines multiple error detection events into one event by hashing, and cross-lane instruction enables error checking via low latency register-level communication. Furthermore, to reduce the performance overhead of software-based recovery, this thesis explores Application-Specific Approximate Recovery (ASAR). ASAR trades-off output quality to reduce the performance penalty of software-based recovery.

Variation affects not just compute but memory components, as well. For memory units, we focus on methods for proactive error prevention. Given the diversity of memory components in use, we focus on emerging Heterogeneous Memory Architectures (HMAs). An HMA consists of multiple memory modules with different performance and reliability characteristics. These differences can be caused by different types of memory modules, different error correcting codes, and effects due to aging. This thesis focuses on methods to place and move data items among memory modules in an HMA system with the goal of reducing the likelihood of encountering an error. Specifically, this work describes two novel data placement techniques: age-aware and vulnerability-aware data placement. The age-aware technique monitors the accumulation of faults in different memory modules as they age, while the vulnerability-aware technique estimates the vulnerability of data in memory to soft errors. The results presented in this work enable practical use of software-based error mitigation solutions for current and future hardware. 

Bio: Manish Gupta is a Ph.D. candidate in the CSE department. His expertise is in GPUs, parallel programming and compilers, and he received a three-year Powell Fellowship when he enrolled at UC San Diego in 2010. Gupta completed his M.S. in 2012, and expects to graduate with his Ph.D. this summer. He did his undergraduate degree at the Indian Institute of Technology, Roorkee (B.Tech. in 2008). Prior to grad school, Gupta worked as a firmware engineer for Qualcomm in Hyderabad, india from 2008 to 2010. He also worked for AMD Research in Austin, TX, for nine months in 2016 as a software engineer, where he worked on reliability versus performance trade-offs for memories. The Ph.D. candidate has over 10 patents pending at the U.S. Patent and Trademark Office, including two patents corresponding to GPU RMT work published at DAC 2017.

 

ALUMNI SURVEY

Take the CSE Alumni Survey. It only takes a few minutes. We'd really like to hear from you!

GIVING OPPORTUNITIES

Image removed.

The Jacobs School of Engineering offers a variety of ways to support the Department of Computer Science and Engineering. As the 2016-'17 academic year gets underway, please consider giving online to the CSE Engineering Tutor Program or the Paul R. Kube Chair of Computer Science.  You can also honor your favorite teacher when you donate to the CSE Teaching Endowment Fund.