Summit Program

For beginning researchers, identifying their research topic can be a daunting and intimidating task. (You mean I need to find the right question and then answer it too?) This talk is part pep talk and part highly practical advice for beginning computer science researchers. I will discuss key aspects of the research process from finding a research topic, to mastering the related work, to keeping up with the onslaught of new research published each year. Some key lessons include how to extract essential research and career advice from a remote mentor of your choice—even if you never have an opportunity to meet them! , how to identify great papers that you can emulate and extend, how to read all (yes all) the papers in your field by reading at varying levels of depth and how to identify insightful criticisms that lead to truly new research directions.

Dr. Jeanna N. Matthews is an Associate Professor of Computer Science at Clarkson University (Potsdam, New York). At Clarkson, she leads several hands-on computing laboratories including the Clarkson Open Source Institute and Clarkson Internet Teaching Laboratory. Students in these labs and in her classes at Clarkson and Cornell have been winners in a number of prestigious computing contests including the 2001, 2002, and 2004 IBM Linux Challenge, the 2005 IBM North American Grid Scholar’s Challenge, the 2005 Unisys Tuxmaster competition, and the 2006 VMware Ultimate Virtual Appliance Challenge.

She has written several popular books including “Running Xen: A Hands-On Guide to the Art of Virtualization” and “Computer Networking: Internet Protocols in Action”. Her research interests include virtualization, cloud computing, computer security, computer networks and operating systems.

She is currently the chair of the ACM Special Interest Group on Operating Systems (SIGOPS), the editor of ACM Operating System Review and a member of the Executive Committee of US-ACM, the U.S. Public Policy Committee of ACM.

She has worked actively on industrial projects with companies including VMware, Intel, IBM, AMD, HP and Greenplum/EMC.

Jeanna received her Ph.D. in Computer Science from the University of California at Berkeley in 1999 and her B.S. in Mathematics and Computer Science from Ohio State University in 1994.

Failures caused by software bugs are widespread in production runs, causing severe losses for end users. Unfortunately, diagnosing production-run failures, especially failures caused by concurrency bugs in multi-threaded software, is challenging. Existing work cannot satisfy privacy, run-time overhead, diagnosis capability, and diagnosis latency requirements all at once.

This talk will present a series of attempts from my group to address the above challenges. Our first attempt, called CCI, uses carefully designed interleaving predicates and random sampling schemes to diagnose a wide variety of concurrency-bug failures with decent overhead. Our second attempt, called PBI, further improves the performance and preserves the diagnosis capability of CCI through a novel use of hardware performance counters. Our final attempt, called LXR, addresses the long diagnosis latency problem of CCI and PBI. Different from CCI and PBI that both obtain run time information through sampling, LXR obtains run-time information through hardware support that maintains recent execution history with negligible overhead.

Shan Lu is the Clare Boothe Luce Assistant Professor of Computer Sciences at the University of Wisconsin—Madison. She earned her Ph.D. at the University of Illinois at Urbana-Champaign in 2008. At the University of Wisconsin, her group works on detecting, diagnosing, and fixing concurrency bugs and performance bugs. Shan Lu won Alfred P. Sloan Research Fellow in 2014, Distinguished Alumni Educator Award from the Department of Computer Science at the University of Illinois in 2013, and NSF Career Award in 2010. Her co-authored papers won the Best Paper Award at USENIX FAST in 2013, ACM-SIGPLAN CACM Research Highlight Nomination in 2011, and IEEE Micro Top Picks in 2006. She serves as the Information Director of ACM-SIGOPS and is currently a proud mom of a one-year-old baby girl.

Machine Intelligence (MI) is the science and engineering of making intelligent machines. The ultimate goal of this field is to make machines reach human-level intelligence. As humans, we take for granted a huge number of capacities and functions that our brains and body are able to seamlessly perform, but are extremely hard to translate to the world of machines. In this talk I plan to discuss the challenges of building machines that can understand, learn, act, and interact in ways that approach human intelligence.

Although machines boast impressive computing speeds and memory, they famously lack many abilities considered basic to human intelligence. To approximate these abilities, machines would have to process sensory data, text and other media and use it to reason, make predictions, and perform actions in the real world. More importantly, they would have to learn from their experiences as well as possess a certain amount of common sense knowledge. For example, a child may never be explicitly told that a coat is used for keeping warm in winter, but she may nonetheless act on such knowledge when it starts to snow. Comparably intelligent behavior might require a machine to have concepts corresponding to winter, snow, warm, and coat, as well as a rich interconnected set of structures that help tie these together and guide its behavior in the context of snow.

Progress toward true machine intelligence would enable a wealth of applications across many domains. This talk will focus on highlighting the main challenges in this area and discuss some of the promising avenues toward new solutions.

Neha is a senior software engineer at Google, where she has worked since 2010. She graduated from Carnegie Mellon University with a Masters in Software Engineering and is currently working with the Machine Intelligence team at Google, contributing to the effort towards understanding common sense knowledge from text and other media. Her main areas of interest include using cognitive methods for language understanding and reasoning.

In her previous role, she worked on improving the quality of news-­related results on Google.com and on Google News. Prior to that, she worked on understanding social interaction signals to improve the quality of advertisements on mobile devices.

God grant me the serenity to accept the things I cannot change, the courage to change the things I can, and the wisdom to know which cool new tool will bring more value than it consumes.

Reveling in the fun of our fast-paced, developer-driven era requires creativity, a twisted sense of delight, and…automation. Genentech's research organizations depend on fast, frequent inception-to-use work. To enable this, we provide a basic tomcat engine, via templates, that is automagically rolled out to the host environments. Our developers adopted automated deployment (Jenkins). Further, we set developers loose with write-access to configs and chase changes using pretty-ok scripts and version control. While evolving this system, (without yet using one of those new-fangled config management systems) we've learned something about how we could do it better. Nothing is perfect, but it's worth doing the imperfect and improving.

Experienced in infrastructure and systems, Meg has worked in large corporations and startups, leading teams in platform integration and software development. Meg spent the past four years working on software infrastructure in Genentech Research and has recently embarked on a new advanture at life360.com

Operating systems sit at a unique place in computing. An OS has always abstracted computing hardware for applications and users alike and does everything from hardware initialization to scheduling algorithms for highly complex NUMA systems, to providing unique and reliable filesystems for massive amounts of data. But an OS is no longer confined to a single computer—as computing infrastructures have exploded, the OS also must deploy, update, and manage massive numbers of systems, virtual machines, and applications. This talk will explore the challenges of creating a truly modern Operating System like Oracle Solaris. It covers a number of specific low-level technology innovations and explores how combining and deeply integrating those capabilities allows us to continue doing the OS's job of abstracting our increasingly complex modern computing infrastructure.

Liane Praza is a Senior Principal Software Engineer and an architect of the Oracle Solaris Operating System. She joined Oracle in 2010 and has been working on the Solaris Operating System for more than 15 years with a variety of technologies, including HA, resource management, service management, and observability. She holds a B.S. in Computer Science from Purdue University.

For the past few years, mobile usage has grown sharply. As an engineer, one needs to always adjust and be aware of latest "trends." Switching focus is not as hard as one might think if you have the right attitude.

Lavinia Petrache has been a software engineer at Facebook since 2011, right after graduating with a computer science degree from a university in Romania. At Facebook, she initially worked on the infrastructure of a spam detection service, but transitioned towards Android development after a year. She is currently working on the Android Photos team.

Release engineering is a relatively new field compared to systems administration. Most companies do release engineering as an afterthought, and many release engineers are wearing multiple hats.

This talk will describe the building blocks of release engineering, the sub-disciplines, and why release engineering is key to delivering high-quality software products. I’ll talk about the skills needed for a top release engineer and make some predictions about what I think is going to happen in the industry over the next few years.

Dinah McNutt is a Release Engineer at Google, Inc. She has 20 years of commercial release engineering experience and has released all types of Unix-based software. She has written for UNIX Review Magazine, Byte, ;login:, and many other publications. She has chaired LISA and is the chair of URES '14.