The Snap! (Build Your Own Blocks) Programming Environment
Dan Garcia
University of California, Berkeley
09/06/2022
Snap! (formerly BYOB) is a visual, drag-and-drop programming language. It is an extended reimplementation of "Scratch" (a project of the Lifelong Kindergarten Group at the MIT Media Lab) that allows users to Build Your Own Blocks. It features first class lists, first class procedures, and continuations. These added capabilities make it suitable for a serious introduction to computer science for high school or college students. This talk (usually offered as a workshop) will introduce the language, and walk users from their first "mobile app" authorable in 90 seconds through the vast array of incredible features the language has to offer.
Beyond Restart: Checkpointing for the Exascale Era
Rebecca Hartman-Baker
User Engagement Group Leader
National Energy Research Scientific Computing Center
09/13/2022
The National Energy Research Scientific Computing Center (NERSC) is the mission high-performance computing center for the United States Department of Energy Office of Science (DOE-SC). NERSC operates cutting-edge, large-scale supercomputers for users performing simulations and data analysis in science areas of interest to DOE-SC. Experimental data analysis, in which data from large scientific instruments such as telescopes or light sources is computationally analyzed, is a growing portion of the compute workload at NERSC. Many of these scientific instruments increasingly require real-time or fast computing turnaround. NERSC's challenge is to support this urgent workload within its existing operations. Checkpoint/restart (C/R) can enable this new workload pattern and address other operational challenges, such as resiliency. In this presentation, we describe our efforts to create a robust offering of checkpointing approaches and to simplify and incentivize their use, in order to optimize the user experience, minimize wasted cycles, and maximize compute-system utilization.
Distance functions on computable graphs
Jennifer Chubb
Associate Professor
University of San Francisco
09/20/2022
An infinite graph---one made of nodes and edges---is computable if there is an algorithm that can decide whether or not a given pair of nodes is connected by an edge. So, for example, the internet is a computable graph which is, for all intents and purposes, infinite. Now, given two nodes on a connected, computable graph, a natural question to ask is, What is the length of the shortest path between them, i.e., the distance between the nodes? Of course, for a connected graph we can always find such a path and determine its length, but the question of finding a shortest path is harder, and is not in general something we can compute for infinite computable graphs.
In this talk, we will see what it means for something to be non-computable via a classic example called the halting problem. Next we will see that it's possible to encode non-computable information into even very simple, computable mathematical objects. Finally, we will see how non-computable information can be encoded into the distance function, the function which outputs the shortest distance between two nodes of a given graph. So, the answer to What is the shortest path between two nodes? Well, we may never know for sure.
This work is joint with Wesley Calvert and Russell Miller.
Deep Learning for Medical Imaging
Geetha Chauhan
AI Practice Head / CTO
SVSG
09/27/2022
The talk covers use cases, special challenges and solutions for Deep Learning for Medical Image Analysis. You will learn about:
- Use cases for Deep Learning in Medical Image Analysis
- Different DNN architectures used for Medical Image Analysis
- Special purpose compute / accelerators for Deep Learning (in the Cloud / On-prem)
- How to parallelize your models for faster training of models and serving for inferenceing.
- Optimization techniques to get the best performance from your cluster (like Kubernetes/ Apache Mesos / Spark)
- How to build an efficient Data Pipeline for Medical Image Analysis using Deep Learning
- Resources to jump start your journey - like public data sets, common models used in Medical Image Analysis
Applications of Nonmaterial Performance to Failure Analysis in High Performance Computing, Side-Channel Vulnerabilities and Facial Hallucination Software
Barry Rountree
Lawrence Livermore National Laboratory
10/04/2022
The Nonmaterial Performance (NMP) frame borrows concepts from Performance Studies, Actor-Network Theory and Vibrant Matter to create a framework for analyzing the performance of software artifacts. In this talk I will be discussing the musical capabilities of the Univac I, loading dynamic libraries on massivesly parallel system, the Platypus side-channel attack, and a machine-learning approach to facial hallucination. Each artifact is examined in terms of the four tenets of NMP: code abstracts, code performs, code acts within a network, and code is vibrant. This work has been done in close collaboration with Dr. William Condee (Ohio University emeritus) and draws from recent work published in The Drama Review, Theater Journal, and an upcoming issue of Digital Humanities Quarterly.
Dr. Rountree holds an BA in Theater from the Ohio University Honors Tutorial College, an MS in System and Network Administration from Florida State University and a PhD in Computer Science from the University of Arizona. He was worked for the past decade as a Computer Scientist at the Center for Applied Scientific Computing at Lawrence Livermore National Laboratory. He has taught two classes on system programming at Sonoma State.