EECS graduate student Ava Jiang Tan (advisor: Sayeef Salahuddin) has won the 2020 Best Paper Award at the 78th Device Research Conference (DRC) for "Reliability of Ferroelectric HfO2-based Memories: From MOS Capacitor to FeFET."  The paper, co-authored by Profs. Salahuddin and Chenming Hu, grad student Yu-Hung Liao, postdoc Jong-Ho Bae, and Li-Chen Wang of MSE, introduces nonvolatile ferroelectric field-effect transistors (FeFETs) which boast impressive programmability and a strong potential for further scalability.  The paper also demonstrates for the first time a systematic, reliable, and rapid method to qualitatively predict the FE endurance of prospective gate stack designs prior to running a full FeFET fabrication process.  Tan works in the Laboratory for Emerging and Exploratory Devices (LEED), and is particularly interested in the architectural potential of nonvolatile ferroelectric CMOS-compatible memories for realizing brain-inspired computing paradigms and energy-efficient hardware for deep learning. The DRC, which is the longest-running device research meeting in the world,  was held in June.

EECS grad student Payam Delgosha is a winner of the IEEE Jack Keil Wolf ISIT Student Paper Award at the 2020 IEEE International Symposium on Information Theory (ISIT), which was held as a Zoomference on June 21 -26, 2020. Payam won the award for his paper "A universal low complexity algorithm for sparse marked graphs" co-authored with his research advisor Venkat Anantharam.  This award recognizes outstanding papers on information theory for which a student is the principal author and presenter. Delgosha earned his B.S. and M.S. degrees from Sharif University of Technology, Iran.  He plans to join the  University of Illinois at Urbana-Champaign as a research assistant professor of computer science in Fall 2020.

EECS Assoc. Prof. Boubacar Kanté, his graduate students Liyi Hsu, Jeongho Ha and Jun-Hee Park, postdoctoral researcher Abdoulaye Ndao, and Prof. Connie Chang-Hasnain, have demonstrated a revolutionary, ultrathin and compact, flat optical lens that spans wavelengths from the visible to the infrared with record-breaking efficiencies.  Their paper, “Octave bandwidth photonic fishnet-achromatic-metalens,” published in Nature Communications, is the first time a photonic system with the entire rainbow has been proposed and demonstrated with efficiencies larger than 70% in the visible-infrared region of the spectrum.  Attempts to make traditional lenses flatter and thinner, so that they can be deployed in increasingly smaller applications, have been hampered by the way that lens curvature and thickness are used to direct light.  The Fishnet-Achromatic-Metalens (FAM) utilizes a complex “fishnet” of tiny, connected waveguides with a gradient in dimensions, which focuses light on a single point on the other side of the lens, regardless of the incident wavelength.  As the world’s thinnest, most efficient, and broadest band, flat lens, its use in applications like solar energy, medical imaging, and virtual reality, is just the beginning.  As Kanté explains, “We have overcome what was regarded as a fundamental roadblock.”  One idea for a possible implementation would be to integrate the miniature lens into microrobots being developed at the Berkeley Sensor & Actuator Center (BSAC).

A paper co-written by EECS alumnus Peter Mattis (B.S. '97) is being presented at the 2020 Association for Computing Machinery (ACM) Special Interest Group on Management of Data (SIGMOD) International Conference on Management of Data this month.  The paper, titled "CockroachDB: The Resilient Geo-Distributed SQL Database," describes a cloud-native, distributed SQL database called CockroachDB, that is designed to store copies of data in multiple locations in order to deliver speedy access.  The database is being developed at Cockroach Labs, a company co-founded in 2015 by a team of former Google employees that included Mattis, who is also the current CTO, and fellow-alumnus Spencer Kimball (CS B.A. '97), currently the company CEO.  Cockroach Labs employs a number of Cal alumni including Ceilia La (CS B.A. '00) and Yahor Yuzefovich (CS B.A. '18).

The EECS department has eleven faculty members who rank among the top 100 most cited computer science & electronics scholars in the world. UC Berkeley ranked #4  in the global list of universities with the highest number of influential scholars in 2020 (35, up from 24 in 2018).  Profs. Michael Jordan, Scott Shenker, Ion Stoica, Jitendra Malik, Trevor Darrell, David Culler, Shankar Sastry, Randy Katz, Alberto Sangiovanni-Vincentelli, Lotfi Zadeh and Dawn Song all ranked in the top 100 with an H-index score of 110 or higher, a measure that reflects the number of influential documents they have authored.   Jordan ranks fourth in the world, with an H-index of 166 and 177,961 citations.  The H-index is computed as the number h of papers receiving at least h citations among the top 6000 scientist profiles in the Google Scholars database. 

Four papers co-authored by EECS faculty (3 of which were co-authored by Prof. Dawn Song) have won Test-of-Time awards at the IEEE Symposium on Security and Privacy today: "Efficient Authentication and Signing of Multicast Streams Over Lossy Channels," co-authored by Song (Ph.D. '02) and the late Prof. Doug Tygar (with Perrig and Canetti) in 2000, "Practical Techniques for Searches on Encrypted Data," co-authored by Song and Prof. David Wagner (with Perrig) in 2000, "Random Key Predistribution Schemes for Sensor Networks," co-authored by Song (with Chan and Perrig) in 2003, and "Outside the Closed World: On Using Machine Learning For Network Intrusion Detection" co-authored by Prof. Vern Paxson (with Sommer) in 2010.    IEEE-SP is considered the premier computer security conference and this four-fold achievement demonstrates Berkeley's preeminence in the field.

Freshly-graduate CS Ph.D. student Daniel J. Fremont (advisor: Sanjit Seshia) has won the Association for Computing Machinery (ACM) Special Interest Group on Embedded Systems (SIGBED) Paul Caspi Memorial Dissertation Award for his thesis on "Algorithmic Improvisation."  The award, which was established in 2013, recognizes outstanding doctoral dissertations that significantly advance the state of the art in the science of embedded systems.  Fremont's thesis proposes a theory of algorithmic improvisation to enable the correct-by-construction synthesis of randomized systems, and explores its applications to safe autonomy.

EECS Prof. Pieter Abbeel and Assistant Prof. Sergey Levine are developing algorithms that enable robots to learn from past experiences — and even from other robots.  They use deep reinforcement learning to bring robots past a crucial threshold in demonstrating human-like intelligence: the ability to independently solve problems and master new tasks in a quicker, more efficient manner.  An article in the Berkeley Engineer delves into the innovations and advances that allow Abbeel and Levine help robots make "good" choices, generalize between tasks, improvise with objects, multi-task, and manage unexpected challenges in the world around them.

EECS Prof. and alumna Dawn Song (Ph.D. '02, advisor: Doug Tygar) and Assistant Prof. Raluca Ada Popa are featured in the cover story for the Spring 2020 issue of the Berkeley Engineer titled "Reinventing Cybersecurity."  Faced with the challenge of protecting users' personal data while recognizing that sharing access to that data "has fueled the modern-day economy" and supports scientific research, Song has proposed a paradigm that involves "controlled use" and an open source approach utilizing a new set of principles based on game theory.  Her lab is creating a platform that applies cryptographic techniques to both machine-learning models and hardware solutions, allowing users to keep their data safe while also making it accessible.  Popa's work focuses on using machine-learning algorithms to keep data encrypted in cloud computing environments instead of just surrounding the data with firewalls.  "Sharing without showing" allows sensitive data to be made available for collaboration without decryption.  This approach is made practical by the creation of a machine-learning training system that is exponentially faster than other approaches. "So instead of training a model in three months, it takes us under three hours.”