EE380 Computer Systems Colloquium

The Future of Computer Architecture (Patterson and Hennessy) [EE380 Computer Systems Colloquium & EE180]

Topic: 
The Future of Computer Architecture (Patterson and Hennessy)
Abstract / Description: 

About the Panel/Q&A (11:30-12:00): The Future of Computer Architecture (Patterson and Hennessy)

 

General Information: This presentation is the last lecture of EE180 for Winter 2017. The lecture is open to the public. The lecture will not be video recorded nor available on the web.

EE380 students and attendees are urged to attend this lecture, but it is optional. It cannot be substituted for one of the ten required EE380 lectures.

 

Date and Time: 
Thursday, March 16, 2017 - 11:30am to 12:00pm
Venue: 
Cubberly Auditorium, School of Education

Service Robots Are Here II [EE380 Computer Systems Colloquium]

Topic: 
Service Robots Are Here II
Abstract / Description: 

After 20 years of predictions that robots will work among us soon, the predictions are finally starting to come true. Investment in robotics is up, enabling start-ups to explore a range of use cases. Decreasing component costs will make it easier to make real business cases for the technology. Mobile robots are beginning to transform the way we serve people in hotels, elder care facilities, hospitals, restaurants, and throughout the service industry, and this trend is accelerating.

This talk is a sequel to my January 2015 talk. I'll discuss what has happened in the Service Robot field over the last two years.

Date and Time: 
Wednesday, March 1, 2017 - 4:15pm
Venue: 
Gates B03

Deep Learning in the Age of Zen, Vega, and Beyond [EE380]

Topic: 
Deep Learning in the Age of Zen, Vega, and Beyond
Abstract / Description: 

Deep Learning and Machine Intelligence is maturing to the point where is it is being deployed to many applications, particularly large data, imaging classification and detection. This talk addresses the challenges of deep learning from a computational challenge perspective and discusses the ways in which new compute platforms of Zen (x86) and Vega (GPU) provide high performance solutions to different training and inference applications. The ROCm software stack completes the support with libraries and framework support for a variety of environments.


ABOUT THE COLLOQUIUM:

See the Colloquium website, http://ee380.stanford.edu, for scheduled speakers, FAQ, and additional information. Stanford and SCPD students can enroll in EE380 for one unit of credit. Anyone is welcome to attend; talks are webcast live and archived for on-demand viewing over the web.

Date and Time: 
Wednesday, January 25, 2017 - 4:30pm
Venue: 
Gates B03

Data For The People [EE380]

Topic: 
Data For The People
Abstract / Description: 

Join us Wednesday at 4:30pm for the Stanford EE Computer Systems Colloquium:
Andreas Weigend will share his journey from starting as a PhD student at Stanford in neural networks, to becoming the Chief Scientist at Amazon, and ultimately writing the book J, that came out last week (Jan 31, 2017) and was displayed on Times Square for free!

Sign up at http://ourdata.com/tour by February 12 (Sunday) to get an electronic copy of the book. Read it. And bring your questions to the talk!

Andreas Weigend just published his book, Data For The People on big data, transparency and what to do about it. As the former Chief Scientist at Amazon, he helped create Amazon's culture of data and innovation.


 The Stanford EE Computer Systems Colloquium (EE380) meets on Wednesdays 4:30-5:45 throughout the academic year. Talks are given before a live audience in Room B03 in the basement of the Gates Computer Science Building on the Stanford Campus. The live talks (and the videos hosted at Stanford and on YouTube) are open to the public.

Date and Time: 
Wednesday, February 15, 2017 - 4:30pm
Venue: 
Gates B03

Beyond Floating Point: Next-Generation Computer Arithmetic [EE380]

Topic: 
Beyond Floating Point: Next-Generation Computer Arithmetic
Abstract / Description: 

A new data type called a "posit" is designed for direct drop-in replacement for IEEE Standard 754 floats. Unlike unum arithmetic, posits do not require interval-type mathematics or variable size operands, and they round if an answer is inexact, much the way floats do. However, they provide compelling advantages over floats, including simpler hardware implementation that scales from as few as two-bit operands to thousands of bits. For any bit width, they have a larger dynamic range, higher accuracy, better closure under arithmetic operations, and simpler exception-handling. For example, posits never overflow to infinity or underflow to zero, and there is no "Not-a-Number" (NaN) value. Posits should take up less space to implement in silicon than an IEEE float of the same size. With fewer gate delays per operation as well as lower silicon footprint, the posit operations per second (POPS) supported by a chip can be significantly higher than the FLOPs using similar hardware resources. GPU accelerators, in particular, could do more arithmetic per watt and per dollar yet deliver superior answer quality.

A series of comprehensive benchmarks compares how many decimals of accuracy can be produced for a set number of bits-per-value, using various number formats. Low-precision posits provide a better solution than "approximate computing" methods that try to tolerate decreases in answer quality. High-precision posits provide better answers (more correct decimals) than floats of the same size, suggesting that in some cases, a 32-bit posit may do a better job than a 64-bit float. In other words, posits beat floats at their own game.


 

The Stanford EE Computer Systems Colloquium (EE380) meets on Wednesdays 4:30-5:45 throughout the academic year. Talks are given before a live audience in Room B03 in the basement of the Gates Computer Science Building on the Stanford Campus. The live talks (and the videos hosted at Stanford and on YouTube) are open to the public.

Date and Time: 
Wednesday, February 1, 2017 - 4:30pm
Venue: 
Gates B03

From chocolate to mice, the darknet to facial expressions, chatbots to humanitarians, and so forth [EE380]

Topic: 
Swiss Computer Systems: From chocolate to mice, the darknet to facial expressions, chatbots to humanitarians, and so forth
Abstract / Description: 

What is beyond the cliché of Switzerland = chocolate ?

  • Is it the numerous and often not well known Swiss contributions to computer science?
  • It is the fact the Switzerland was the birthplace of the world wide web?
  • Is it the "porous" system, largely unbound by geographies, which encourages transdisciplinarity?
  • Is it the century-old tradition of neutrality and humanitarian engagement challenged by the ubiquity of communication devices and social networks?
  • Is it the power of artists thinking about the societal impact of IT?

In this talk I'll use several examples from here and there to discover similarities and differences, and how they will shape the future.


 

The Stanford EE Computer Systems Colloquium (EE380) meets on Wednesdays 4:30-5:45 throughout the academic year. Talks are given before a live audience in Room B03 in the basement of the Gates Computer Science Building on the Stanford Campus. The live talks (and the videos hosted at Stanford and on YouTube) are open to the public.

Date and Time: 
Wednesday, January 18, 2017 - 4:30pm
Venue: 
Gates B03

EE380 Computer Systems Colloquium: Charting a Cybersecurity Path for the Next Administration

Topic: 
Charting a Cybersecurity Path for the Next Administration: Report of the President's Commission on Enhancing National Cybersecurity
Abstract / Description: 

In February 2016, President Obama established the Commission on Enhancing National Cybersecurity, and charged it with developing a cybersecurity roadmap for the next Administration. On Friday, December 2, the Commission released its final report. As one of the 12 commissioners responsible for the report, I will brief the major themes and recommendations in the report, provide some personal reflections, and answer questions as best I can.

Date and Time: 
Wednesday, December 7, 2016 - 4:30pm to 5:30pm
Venue: 
Gates B03

EE380: The Time-Less Datacenter

Topic: 
The Time-Less Datacenter
Abstract / Description: 

We describe, and demonstrate, a novel foundation for datacenter communication: a new "event based" protocol that can dispense with the need for conventional heartbeats and timeouts at the network layer --paving a new path for efficient recovery for distributed algorithms as they scale. We then show how this can be composed into arbitrary graph-based distributed communications for application infrastructures.

A modern datacenter has many thousands of servers connected to each other through many hundreds of switches. The switches are usually configured into a spanning tree with the servers at the leaves. While this approach simplifies routing, it has some serious shortcomings. For example, servers don't know when an interior link has failed, so they use timeouts as a way of guessing that they need to fail over to a new path. Since failovers generate lots of messages, they often result in other servers timing out. Sometimes, several minutes elapse before the system quiesces and normal operations can resume. These failover and latency storms would be tolerable if they were rare, but they have been observed to occur several times a day in modern datacenters.

The Earth Computing Network Fabric (ECNF) takes a different approach. An ECNF segment within a datacenter has no switches within its own fabric. Instead, each cell combines the compute functions of a server with the routing functions of a switch. Each cell has multiple ports (7±2), and each port of a cell is directly connected to a port of another cell via a link. Because the link is a dedicated channel between exactly two cells, we can use the Earth Computing Link Protocol (ECLP) instead of standard protocols, such as Ethernet or TCP/IP.

In this talk, we'll explain the problems with the way datacenters are built today and show how the Earth Computing design avoids many difficult problems while providing additional functionality.

If you can, attend this talk live. Following the formal presentation, we are planning to demonstrate and discuss "interesting things" off camera. For example, questions on the proprietary nature of the Implementation will be addressed only during the extended session when the camera is turned off.

The following background may be helpful to computer science students unfamiliar with the nature of time in physics:

Date and Time: 
Wednesday, November 16, 2016 - 4:30pm to 5:30pm
Venue: 
Gates B03

EE380 Computer Systems Colloquium

Topic: 
Runway: a new tool for distributed systems design
Abstract / Description: 

Distributed systems are notoriously difficult to get right. We're constantly improving the frameworks we use and the way we test production code, yet we rarely invest in tools for system design. The best design tools we have are still low-tech, like whiteboards and back-of-the-envelope calculations. Runway is a new tool to help design distributed and concurrent systems. It combines specification, simulation, visualization, and model checking into one tool, so you can write a single system model and do a bunch of useful things with it: the same model can help you evaluate a design's correctness, availability, and performance while also helping others learn how the design works.

Date and Time: 
Wednesday, November 2, 2016 - 4:30pm to 5:30pm
Venue: 
Gates B03

Pages

Subscribe to RSS - EE380 Computer Systems Colloquium