MLPerf: SPEC for ML

David Patterson Deep Learning, News, Open Source, Optimization, Reinforcement Learning, Systems, Uncategorized 0 Comments

The RISE Lab at UC Berkeley today joins Baidu, Google, Harvard University, and Stanford University to announce a new benchmark suite for machine learning called MLPerf at the O’Reilly AI conference in New York City (see https://mlperf.org/). The MLPerf effort aims to build a common set of benchmarks that enables the machine learning (ML) field to measure system performance eventually for both training and inference from mobile devices to cloud services. We believe that a widely accepted benchmark suite will benefit the entire community, including researchers, developers, builders of machine learning frameworks, cloud service providers, hardware manufacturers, application providers, and end users. Historical Inspiration. We are motivated in part by the System Performance Evaluation Cooperative (SPEC) benchmark for general-purpose computing that drove rapid, …

Prof. David Patterson, one of the founders of RISELab, has been awarded the Turing Award!

Jon Kuroda News 0 Comments

We are proud to announce that Professor David Patterson, one of the RISELab founders, is a recipient of the 2017 ACM A.M. Turing Award, the annual prize given by the Association for Computing Machinery (ACM) to “an individual selected for contributions of a technical nature made to the computing community,” and often referred to as the “Nobel Prize of Computing”. He shares the award with John L. Hennessy, former President of Stanford University, for their invention of the RISC processor. For more details, please read the official ACM announcement. Congratulations to Prof. Patterson on joining the highest ranks of computer science luminaries!

Michael Jordan named as Plenary Lecturer at the 2018 International Congress of Mathematicians (ICM) in Rio de Janeiro!

Angie Goodwin News 0 Comments

EECS faculty continues to make big news in the outside world! Michael Jordan, the Pehong Chen Distinguished Professor in EECS and the Department of Statistics, has been named a Plenary Lecturer at the 2018 International Congress of Mathematicians (ICM) in Rio de Janeiro this August. This is an honor that has been bestowed upon a very small handful of computer scientists over the 100-year history of the ICM. (For the uninitiated, a plenary lecturer is one who gives his or her talk to all attendees of a conference at once. It’s a pretty huge deal.) Here is the list of plenary speakers at the ICM 2018 Conference. The title of Prof. Jordan’s speech will be “Dynamical, Symplectic and Stochastic Perspectives on Gradient-Based Optimization” …

NSF Expeditions Proposal Awarded to RISELab!

Angie Goodwin News 0 Comments

The UC Berkeley RISELab has been awarded the prestigious NSF Expedition Award by the National Science Foundation (NSF).  The expedition award, which provides $10 million in funding over five years, is awarded to only three labs in the country. This funding will enable the RISELab to make fundamental advances in the theory and design of real-time, intelligent, secure, and explainable systems.  This research has the potential to transform the next generation of systems that interact with the real-world, including autonomous vehicles, medical robots, privacy-preserving information exchanges, fraud detection systems, and the power grid. Read the full NSF press release and Berkeley press release for more details.

Pictures from RISE Winter 2018 retreat!

Boban Zarkovich News

Our biggest retreat so far (182 registered attendees!) happened January 10 – 12, 2018, at Monterey Tides hotel. We had lots of productive interactions (presentations, meeting, poster sessions), but also managed to have some fun kayaking at the Elkhorn Slough! You can see the photos on our Facebook page.

Incubator launches AI-focused accelerator to help startups out of UC Berkeley, RISELab faculty involved

Boban Zarkovich News

Berkeley-based startup incubator and venture capital fund The House is launching an AI-specific initiative backed by Google that is meant to take advantage of the technical and academic expertise in the field coming out of UC Berkeley. The faculty involved in the program include Databricks co-founder and UC Berkeley professor Ion Stoica, Berkeley Artificial Intelligence Research lab co-director Trevor Darrell and UC Berkeley professor Michael Jordan, one of the most influential computer scientists working in AI. You can read the whole article here.

November 2017 Newsletter Published!

Angie Goodwin News

The third installment of the RISELab newsletter went out on November 15. You can read all about what’s been going on here. If you would like to be added to our mailing list, simply fill out this form. Happy reading!

RISELab to cooperate with Alibaba DAMO Academy

Boban Zarkovich News

Alibaba Group launched a massive global research program, “Alibaba DAMO Academy”, by committing $15B in funding. The Academy aims to increase technological collaboration worldwide, advance the development of cutting-edge technology, and strive to make the world more inclusive by narrowing the technology gap. The Academy will cooperate with the University of California, Berkeley through its RISE Lab on areas such as secured real-time computing. For more information, please read this article.

Prof. Popa gave a keynote talk at the GE Industrial Machine Learning Workshop 2017

Boban Zarkovich News

RISELab’s Raluca Ada Popa gave a keynote talk on security for machine learning systems at the GE Industrial Machine Learning Workshop 2017. https://wise.io/imlw17/ Her talk discussed security threats in deploying machine learning systems as well as three promising approaches to address some of these: secure multi-party computation, differential privacy and hardware enclaves.

RISELab and UC Berkeley’s new Foundations of Data Analysis (FODA) Institute

Michael Mahoney News

We are pleased to announce the involvement of RISELab faculty with the new Foundations of Data Analysis (FODA) Institute. The FODA Institute is funded as part of the NSF TRIPODS program, and it is designed to bring together core research communities in theoretical statistics, applied mathematics, and theoretical computer science to address foundational questions at the heart of data science. In addition to RISELab faculty Michael Mahoney and Mike Jordan, also involved with this FODA/TRIPODS effort are statistics faculty Bin Yu and Fernando Perez and EECS/Simons faculty Dick Karp. We look forward to the fruitful interactions between theory, implementations, and real-time applications.   For more details, here is the link to the press release: https://data.berkeley.edu/news/berkeley-defining-next-academic-frontier-two-nsf-awards-uc-berkeley-set-stage-new-era-data

RISELab faculty publishes paper on how to build more secure, faster AI systems

Boban Zarkovich News

The RISELab faculty—who cover the fields of machine learning, systems, security, and hardware—have just published a vision paper on systems challenges for AI. They argue that future AI systems must make timely and safe decisions in unpredictable environments, be robust against sophisticated adversaries, and process ever increasing amounts of data across organizations and individuals without compromising confidentiality. These challenges will be exacerbated by the end of the Moore’s Law, which will constrain the amount of data these technologies can store and process. They propose several open research directions in systems, computer architecture, and security that can address these challenges and help unlock AI’s potential to improve lives and society. Link to the press release: https://news.berkeley.edu/2017/10/16/berkeley-experts-on-how-to-build-faster-safer-and-more-secure-ai-systems/

ActiveClean featured in “the morning paper”

Ion Stoica News

ActiveClean has been featured in today’s “the morning paper“. The ActiveClean project aims to develop tools and algorithms to address one of the key steps in model training pipelines: handle dirty or inconsistent data including extracting structure, imputing missing values, and handling incorrect data.

PyWren wins Best Vision Paper Award at SOCC’17

Ion Stoica News

PyWren won the Best Vision paper Award at SOCC’17. PyWren is a new paralle computation engine that drmatically lowers the barrier for scientists to use public cloud for massively parallel worklods, by obviating the need for complex cluster management. This is a joint work between RISELab and Berkeley Center for Computational Imaging.

Second RISELab Newsletter!

Boban Zarkovich News

In case you missed it (it went out on August 1, 2017), you can still read a web copy here. If you’d like to be added to our mailing list, please fill out this form.

Announcing Ground v0.1

Vikram Sreekanti blog, Ground, News, Open Source, Projects, Systems

We’re excited to be releasing v0.1 of the Ground project! Ground is a data context service. It is a central repository for all the information surrounding the use of data in an organization. Ground concerns itself with what data an organization has, where that data is, who (both human beings and software systems) is touching that data, and how that data is being modified and described. Above all, Ground aims to be an open-source, vendor neutral system that provides users an unopinionated metamodel and set of APIs that allow them to think about and interact with data context generated in their organization. Ground has many use cases, but we’re focused on two specific ones at present: Data Inventory: large organizations …

RISELab and the 5G Innovators Initiative (5GI2)

Randy Katz News, Uncategorized

5G, also known as Fifth Generation Mobile Networks, is an emerging global telecommunication system designed for the next generation of significantly higher wireless data bandwidths to support a variety of consumer, commercial, and industrial applications. On promise are data rates of 10-100 mbps for tens of thousands of simultaneous users in the metropolitan area, with 1 gbps indoors and connectivity for hundreds of thousands of simultaneously connected sensors. As important as these enhanced bandwidths will be the software extensibility and configurability of the 5G network, making it possible to partition and customize network bandwidth and services for a variety of site- and area-specific applications to support diverse devices at the network edge. RISELab and our industrial sponsors Ericsson, Intel, and …

Shivaram Venkataraman has won the 2016-2017 “Demetri Angelakos Memorial” Achievement Award

Ion Stoica News

Shivaram Venkataraman has received the 2016-2017 “Demetri Angelakos Memorial” Achievement Award  who recognizes students that “in addition to conducting research, unselfishly take the time to help colleagues beyond the normal cooperation existing between fellow students“. There is hard to imagine a more deserving recipient than Shivaram. During his PhD, Shivaram has been generous with his time and sharing credit to a fault. He has helped his peers in every imaginable way; after six years his colleagues still have to hear him saying “no” when asked for help. He has been a trusted sounding board for other graduate students (and even faculty) when it comes to feedback on their research, talks, and papers. Shivaram has been, without exaggeration, the nexus of knowledge …

Wenting Zheng is Awarded the 2017-18 IBM PhD Fellowship

Joseph Gonzalez Crypto, News, Security

Wenting Zheng was awarded the prestigious IBM PhD Fellowship for her work on  security and distributed systems. Wenting is actively studying new methods for scalable secure analytics, multi-party computation for machine learning, and distributed zero knowledge proofs.  The IBM Ph.D. fellowship is an “intensely competitive worldwide program that honors exceptional Ph.D. students who have an interest in solving problems that are important to IBM and fundamental to innovation in many academic disciplines and areas of study.” Only 50 fellowships are awarded worldwide annually.

Join RISELab at NSDI’17

Boban Zarkovich News

Join RISELab at NSDI’17 on March 27th through the 29th in Boston.  There we will be presenting the Opaque and Clipper systems. Please see the below links to read the final versions of the papers and try pre-release versions of the software and check out slides and videos from their previous talks. Opaque system: The paper is already available here Software is available here Slides and video from a previous talk here Clipper system: The paper is available here Software is available here Slides and video from a previous talk here