conference-talk

Experiences Developing a Fortran Code Transformer

Date and Time: 
2014 April 8th @ 8:30am
Location: 
CG1 Auditorium
Speaker: 
Paul Madden

The Parallel Preprocessor (ppp), a component of the the Scalable Modeling System (SMS, a parallelization toolkit for Fortran developed at NOAA), translates Fortran code with user-supplied directives into a parallel form, based on MPI, ready for compilation. Initially developed in the early '90s, ppp until recently had less than robust support for Fortran 90 (and beyond), and was in need of an overhaul.

Speaker Description: 

I am a software engineer in the Advanced Computing Section of NOAA's Earth System Research Laboratory, Global Systems Division, via CU Boulder's Cooperative Institute for Research in Environmental Sciences. My work mainly involves automatic parallelization via language translation, regression-testing infrastructure, build/run automation, and tool support (revision control, profiling, etc.) for scientific-model development.

Event Category:

Testing made easy

Date and Time: 
2014 April 8th @ 1:30pm
Location: 
CG1 Auditorium
Speaker: 
Robert McLay

Incorporating testing as part of a software development project, while always seen as a noble goal, is typically not done in many projects, especially in academic ones. This talk will cover some easy ways to make testing part of projects that do numerical work. Most programs either grow or die. New requirement come in and require refactoring. Testing can make this process easier and even enjoyable.

Speaker Description: 

Doctor Robert McLay received bachelors and masters degree from the Massachusetts Institute of Technology and his Ph.D in Engineering Mechanics from The University of Texas at Austin. His research include C++ software development, regression testing, and software tools, all related to large parallel numerical simulation codes. In particular, he has done work in parallel finite-element programs solving incompressible fluid flow and heat transfer.

His interest in software tools and support of HPC programming environments has lead to his development of Lmod, a modern replacement for Environment Modules system. Lmod's major advantages are protect all users from loading incompatible software without hindering experts. This work as lead to an interest in tracking the software usage through the module system.

Event Category:

Code Testing in a Distributed Environment Lesson Learned from a NCAR-Academia Project

Date and Time: 
2014 April 8th @ 2:00pm
Location: 
CG1 Auditorium
Speaker: 
Guido Cervone

Code development in a distributed environment is often associated with challenges due to the geographical constraints, different systems and software development practices. In an academic environment, these problems are paired with a relatively high turn-around of developers with different backgrounds and capabilities. This talk addresses lessons learned relative to code testing of the Analog Ensemble (AnEn) methodology, performed at NCAR and at the Pennsylvania State University.

Speaker Description: 

I am an Associate Professor at the Department of Geography, Institute for CyberScience, GeoVISTA Center The Pennsylvania State University, University Park, PA

My fields of expertise are geoinformatics, machine learning and remote sensing. My research focuses on the development and application of computational algorithms for the analysis of spatio-temporal remote sensing, numerical modeling and social media “Big Data” related to man-made, technological and environmental hazards. I operate a satellite receiving station for NOAA POES satellites. I received research funding from ONR, DOT, NASA, Italian Ministry of Research and Education, Draper Labs, Stormcenter Communication.

I am a member of the advisory committee of the United National Environmental Programme, division of Disasters and Early Warning Assessment. I am also affiliated faculty in the Research Application Laboratory (RAL) at the National Center for Atmospheric Research (NCAR). In 2013 I received the “Medaglia di Rappresentanza” from the President of the Italian Republic for my work related to the Fukushima crisis. I received the 2013 ISNAAF award. I co-chaired the 2010 SIGSPATIAL Data Mining for Geoinformatics (DMG-10) workshop. I served as the program co-chair for the 2008 and 2009 IEEE International Conference on Data Mining (ICDM) Spatial and Spatio-Temporal Data Mining (SSTDM) workshop.

I authored two edited books, over forty fully refereed articles relative to data mining, remote sensing and environmental hazards. In 2010, I was awarded a US patent for an anomaly detection algorithm. My research on natural hazards was featured on TV news and newspapers, on general interest magazines such as National Geographic, and on international magazines.

Event Category:

Getting the most out of RT: Library and REST API Development

Date and Time: 
2014 April 8th @ 4:15pm
Location: 
CG1 Auditorium
Speaker: 
Carrie Arnold

Software projects, both large and small, benefit from relying on an issue tracking system to manage and maintain a central and up to date list of project features and bugs. Ticketing systems have become a fundamental component in large eScience projects and advanced computing institutes to keep track of user issues and requests. Projects such as XSEDE and Blue Waters and advanced computing centers such as TACC, use Request Tracker (RT) as a help desk ticketing system for issue tracking.

Speaker Description: 

Carrie Arnold is a developer in the Web and Mobile Applications group at the Texas Advanced Computing Center (TACC) at The University of Texas at Austin. Her diverse programming background has her working on multiple projects spanning development in Drupal, Java, and Liferay portal development.

Event Category:

Productivity-oriented software design for geoscientific modelling

Date and Time: 
2014 April 7th @ 2:00pm
Location: 
CG1 Auditorium
Speaker: 
Dorota Jarecka

Researchers' productivity in computational science is increasingly determined by software quality. From the users' perspective it means: ease of use, robustness, result reproducibility. From the developers' perspective it means extendability and maintainability. In our community, many researchers are users and developers alike.

Speaker Description: 

Dorota Jarecka is an Assistant Professor at the University of Warsaw (Warsaw, Poland) and a Visitor within the Mesoscale and Microscale Meteorology Division of NCAR. She received her PhD in Physics from University of Warsaw in September 2012. Her research interests include cloud microphysics, atmospheric numerical simulations and scientific computing with Python. She just started a new project dedicated to develop and test microphysical schemes in numerical models.

Event Category:

DevOps in the Data Center

Date and Time: 
2014 April 7th @ 8:30am
Location: 
CG1 Auditorium
Speaker: 
Rion Dooley

The Agave Developer APIs are a hosted, cloud-based collection of REST APIs bringing HPC, HTC, and Big Data to the web. One of the challenges in providing hosted software is designing and maturing our devops infrastructure to give us, at the same time, highly accurate monitoring, useful logging, continuous integration and continuous deployment, and a scalable infrastructure that can be quickly stood up or torn down based on demand. This talk focuses on the successes and learning moments we had trying to "get stuff working" and "keep stuff working" reliably.

Speaker Description: 

Rion Dooley is a research associate at the Texas Advanced Computing where he leads the Web and Cloud Services group. He earned his Ph.D. in CS from LSU in 2004 with the support of a Board of Regents Fellowship. He worked at LSU's Center of Computation and Technology for 2 years before moving to UT. Past projects include the development of science gateways such as GridChem, the TeraGrid mobile user portal, and the XSEDE user portal. Rion is currently the lead architect of the Agave API, and co-PI of the Distributed Web Security for Science Gateways project. He is also a senior participant in the Science Gateway Institute planning project. Rion's primary research interests include distributed systems, cloud infrastructure, and data management.

Event Category:

Designing and Building Awesome REST APIs That Get Used

Date and Time: 
2014 April 8th @ 3:45pm
Location: 
CG1 Auditorium
Speaker: 
Rion Dooley

Everyone loves a great app, but how many of us have stopped to consider the role the supporting APIs played in making the app great. Your choice of APIs can make or break a project. And, as developers and service providers, the design and development of your own APIs can make or break the projects of the people attempting to leverage your resources. In this talk, we walk through what goes into designing and building an awesome REST API that gets used.

Speaker Description: 

Rion Dooley is a research associate at the Texas Advanced Computing where he leads the Web and Cloud Services group. He earned his Ph.D. in CS from LSU in 2004 with the support of a Board of Regents Fellowship. He worked at LSU's Center of Computation and Technology for 2 years before moving to UT. Past projects include the development of science gateways such as GridChem, the TeraGrid mobile user portal, and the XSEDE user portal. Rion is currently the lead architect of the Agave API, and co-PI of the Distributed Web Security for Science Gateways project. He is also a senior participant in the Science Gateway Institute planning project. Rion's primary research interests include distributed systems, cloud infrastructure, and data management.

Event Category:

Agent Based Modeling in HPC

Date and Time: 
2014 April 8th @ 10:15am
Location: 
CG1 Auditorium
Speaker: 
Mike Page

In this talk, I will offer a definition of Agent Based Modeling (ABM), review a list of software tools used to conduct ABM, discuss some ABM application areas and present some results from the field. I will also show how one ABM tool (FLAME) has made the transition from serial execution to a large-scale, distributed memory hardware environment. This will lead to a call for ABM studies of large populations in social and computational sciences.

Speaker Description: 

Mike Page is currently an HPC Software Analyst seeking contract opportunities under the name 'Theory and Practice’. His career in HPC has spanned from the time he worked for Cray Research, Inc. He was a Software Engineer in NCAR’s CISL and then RAL from 2003 to 2012.

Theory and Practice offers HPC application support and optimization services.

Mike gave his first presentation centered on agent-based modeling at a UN conference on the Human Dimensions of Global Environmental Change in Potsdam, Germany in 2005 near the building where the Michelson-Morley experiments on the speed of light were performed.

Event Category:

OAuth in the Agave Platform

Date and Time: 
2014 April 8th @ 3:15pm
Location: 
CG1 Auditorium
Speaker: 
Joe Stubbs

The OAuth specification provides a powerful and secure framework for delegating access to sensitive data and resources over the web. Version 2 was published in October of 2012 and has seen wide adoption from industry leaders like Google, Facebook and Twitter. In this talk we will present an overview of OAuth 2 and its uses. We'll then discuss how we built support for OAuth2 into Agave, our 'science as a service' platform. In concert with other technologies like MyProxy and GSI security, we'll show how OAuth and Agave can significantly simplify the challenges of bringing HPC to the web.

Speaker Description: 

Joe Stubbs earned a PhD in Mathematics from the University of Michigan. Since then he has been at the University of Texas where he has focused on building infrastructure software in various contexts. He is currently a research scientist at TACC where he primarily works on the Agave "science as a service" platform, enabling the next generation of science gateways to harness petascale HPC over the web.

Event Category:

Python Tools for Parallel Analysis of Extremely Large GCM Output

Date and Time: 
2014 April 11th @ 1pm
Location: 
CG - room TBD
Speaker: 
Ryan Abernathey

Ocean General Circulation Models (GCMs) are resolving finer and finer scales, meaning that the size of the computational domain is growing rapidly. While the parallel achitecture of GCM codes themselves scales well, the typical tools we use to analyze the output (where the actual science happens) do not. I will report on a toolkit I am developing for the parallel analysis of extremely large global GCM simulations (https://github.com/rabernat/MITgcm_parallel_analysis).

Speaker Description: 

Assistant Professor of Earth and Environmental Sciences @ Columbia University / Lamont Doherty Earth Observatory

prevously: Postdoc at Scripps Institution of Oceanographu Ph.D. in Climate Physics and Chemistry at MIT

interests: Global ocean circulation, mesoscale eddy dynamics, transport and mixing in turbulent flows

Event Category:

Pages

Subscribe to conference-talk