Graduate

High reliability and availability is a requirement for most technical systems. Reliability and availability assurance methods based on probabilistic models is the topic addressed in this talk. Non-state-space solution methods are often used to solve models based on reliability block diagrams, fault trees and reliability graphs. Relatively efficient algorithms are known to handle systems with hundreds of components and have been implemented in many software packages. Nevertheless, many practical problems cannot be handled by such algorithms. Bounding algorithms are then used in such cases as was done for a major subsystem of Boeing 787. Non-state-space methods derive their efficiency from the independence assumption that is often violated in practice. State space methods based on Markov chains, stochastic Petri nets, semi-Markov and Markov regenerative processes can be used to model various kinds of dependencies among system components. However, the resulting state space explosion severely restricts the size of the problem that can be solved. Hierarchical and fixed-point iterative methods provide a scalable alternative that combines the strengths of state space and non-state-space methods and have been extensively used to solve real-life problems. We will take a journey through these model types via interesting real-world examples chosen from IBM, Cisco, Sun Microsystems, and Boeing. These methods and applications are fully described in a recently completed book: Reliability and Availability Engineering: Modeling, Analysis and Applications, Cambridge University Press, 2017.

Quantization plays a critical role in digital signal processing systems. Quantizers are typically designed to obtain an accurate digital representation of the input signal, operating independently of the system task, and are commonly implemented using serial scalar analog-to-digital converters (ADCs). This talk is concerned with hardware-limited task-based quantization, where a system utilizing a serial scalar ADC is designed to provide a suitable representation in order to allow the recovery of a parameter vector underlying the input signal. We propose hardware-limited task-based quantization systems for a fixed and finite quantization resolution, and characterize their achievable distortion. Our results illustrate the benefits of properly taking into account the underlying task in the design of the quantization scheme.

The Keynote speaker at this year's Japan – U.S. Innovation Awards Symposium is artificial intelligence and robotics sensation Dr. Yoky Matsuoka, CTO of Nest.

Dr. Matsuoka grew up in Japan assuming she would become a professional tennis player, but instead earned a BS degree from UC Berkeley and a PhD from MIT. Subsequently, she became a professor at Carnegie Mellon University, then the University of Washington, developing robotic devices for rehabilitating and assisting the human body and brain. This work led her to receive the MacArthur award.

In late 2009, Dr. Matsuoka joined Google[x] as one of the three founding members before taking her position as vice president of technology at Nest. She was named CEO of a small startup, Quanttus, in 2015, and a year later joined Apple's Health Group before returning to Nest in 2017 as CTO.

Please come to hear from one of Silicon Valley's top technologists and to network with people who share an interest in innovation connections between the Bay Area and Japan!

Stanford ID holders, please email briana.burrows@stanford from your Stanford.edu email address for a complimentary ticket.

Japanese companies have been coming to Silicon Valley for many decades, but in recent years that has turned into a flood. More than 100 Japanese companies have opened up new operations in the Bay Area over just the past two years, including giants like Toyota (TRI) and FujiFilm (Open Innovation Lab). What has sparked this renewed Japanese interest in Silicon Valley? Will Japanese firms fare better this time as they struggle to take full advantage of the Bay Area's entrepreneurial ecosystem? What lessons can they learn from their past missteps and what actions can they take this time to succeed?

Few are better qualified to talk about these issues than Masa Ishii, who has been a direct participant in Japan's engagement in Silicon Valley for more than 40 years, as a student, consultant, investor and academic observer. He will discuss the challenges that Japanese companies face in getting the most out of their Bay Area innovation networks, provide some examples of failures and successes, and present his recommendations for approaches that Japanese corporations can take to better ensure successful business outcomes. Please come join this fascinating discussion that is critical to Japan's economic success.

Python is a dynamically typed language, and some of its appeal derives from this. Nevertheless, especially for large code bases, it would be nice if a compiler could find type errors before the code is even run. Optional static type checking promises exactly this, and over the past four years we have successfully introduced this feature into Python 3. This talk introduces the type system we've adopted and the syntax used for type annotations, some tips on how to get started with a large existing code base, and our experience using the 'mypy' type checker at Dropbox. The entire system is open source, and has also been adopted by other companies such as Lyft, Quora and Facebook.

You are cordially invited to come see the demos and posters presented by the EE267 students!

Spring quarter at Stanford is coming to an end and with that our latest offering of EE267 "Virtual Reality" (more info). The students have been working hard on their course projects for the last few weeks and will present demos and posters of innovative VR technology and applications they have been working on to the public – come check them out!

In the past decade, CDSC has been exploring customizable computing, which emphasizes extensive use of customized accelerators on programmable fabrics for much greater performance and energy efficiency. With Intel's $17B acquisition of Altera in 2015 and Amazon's introduction of FPGAs in its AWS public cloud in 2017, customizable computing is going from advanced research into mainstream computing.

Although the performance and energy efficiency benefits have been clearly demonstrated, a significant challenges, however, is the efficient design and implementation of various acceleration on FPGAs, which is a barrier to many software programmers. In this talk, I shall talk about our effort on developing an automated compilation flow from high-level programming languages to FPGAs. I start with a quick review of our early work on high-level synthesis. Then, I shall present our recent effort on source-code level transformation and optimization for customizable computing, including support of high-level domain-specific languages (DSL) for deep learning (with Caffe or TensorFlow), imaging processing (with Halide), and big-data processing (with Spark), and support automated compilation to customized microarchitecture templates, such as systolic arrays, stencils, and CPP (composable parallel and pipelined).

This talk provides an overview of Siemens Corporate Technology's recent research on new control functions for future power systems. Three different topics are discussed: (a) adaptive power oscillation damping optimization to increase the stability reserve of power systems, (b) robust power flow optimization to increase power system resilience to volatile generation, and (c) new research challenges for autonomous microgrids that provide autonomous operation and plug-and-produce capabilities.

Visualizing the dynamic movements and interactions between biomolecules remains a challenge, motivating the development of new optical technology and computational algorithms for imaging at the nanoscale. We have built two technologies for multidimensional imaging of single molecules (SMs): the Tri-spot point spread function (PSF) and the Robust Statistical Estimation (RoSE) algorithm. The Tri-spot PSF measures each second moment of SM orientation with near-uniform sensitivity, thereby capturing the orientation and rotational diffusion of SMs using just one camera frame. For 3D imaging, we developed RoSE to minimize the vectorial localization errors in super-resolution microscopy that result from both the structure of the sample and the PSF itself. By estimating the likelihood of a blinking event to be present in each imaging frame, RoSE localizes molecules accurately and minimizes false localizations even when images overlap.

This talk introduces a "differential" approach to information theory. In contrast to the more traditional "elemental" approach, in which we work to understand communication networks by studying the behavior of their elements in isolation, the differential approach works to understand the impact components can have on the larger networks in which they are employed. Results achieved through this differential viewpoint highlight some startling facts about network communications -- including both opportunities where even very small changes to a communication network can have a big impact on network performance and vulnerabilities where small failures can cause big harm.