Home » conference-talk

conference-talk

Introducing CloudyCluster, Self Service Elastic HPC in the Cloud

Location: 
CG Auditorium
Speaker: 
Boyd Wilson

Interested in learning how to provide a familiar HPC environment in the cloud?  In this session, we will demonstrate the flexibility and scalability of CloudyCluster and its ability to provide near seamless execution of on-premises HPC jobs in th

Speaker Description: 

President and CEO of Omnibond.

Event Category:

Developing OpenOA: An Open Source Framework for the Operational Analysis of Wind Plants

Location: 
CG Auditorium
Speaker: 
Jordan Perr-Sauer

Operational data from wind power plants is used across the wind industry to perform performance and reliability analysis of these electricity generation assets. As part of an industry-wide, NREL-led benchmarking project, the need for an open source implementation of a standard type of analysis was identified. In this talk, I will describe the different phases of our development process within the national laboratory context, and share the lessons and experience we've gained along the way.

Speaker Description: 

Jordan Perr-Sauer is a data science researcher in the computational sciences center at NREL. He has a background in applied mathematics and software development and wants to use software to answer interesting scientific questions. Jordan has a passion for improving scientific software, sitting on the steering committee for the United States Research Software Engineers (US-RSE), which is a new organization meant to promote and improve the practice of software engineering in research organizations. Jordan has been a participant in the NREL-led WP3 Benchmarking initiative for the past two years, serving in a leadership role in the development and promotion of OpenOA.

Event Category:

Parallel Analog Ensemble -- The Power of Weather Analogs

Location: 
CG Auditorium
Speaker: 
Weiming Hu

The Analog Ensemble (AnEn) is a technique used to generate ensemble forecasts from a single simulation of a deterministic numerical weather prediction (NWP) model. In contrast to relying on multi-model or multi-realization approaches, AnEn generates ensemble forecasts by searching through a historical repository for the most similar past weather forecasts and selecting ensemble members from observations that are associated with the most similar historical forecasts.

Speaker Description: 

Weiming Hu is a Ph.D. candidate at Penn State University in the Dept. of Geography focusing on computational algorithms and geographic information science.

Event Category:

Facilitating Discovery of Scientific Notebooks: Applying Schema.org and CodeMeta Metadata to Jupyter Notebooks

Location: 
CG Auditorium
Speaker: 
Keith Maull

Jupyter Notebooks and related computational tools are becoming commonplace
in scientific analyses, research explorations, publications, educational
tutorials and nearly anywhere computational narratives are used.
The rapid explosion of these notebooks has resulted in the availability of
millions of open notebooks on platforms like Github, but a corresponding
lack of both metadata and metadata standards has significantly hindered the
widespread discovery of all but the most popular of these important class

Speaker Description: 

Keith is a software engineer and data scientist at the NCAR Library specializing in scholarly metrics and digital scholarship from data to software to computational narratives. Keith joined the NCAR library in 2013 after completing his PhD in computer science from CU/Boulder, and has been passionately involved in research and development projects at the Library, including bibliometrics and scholarly metrics initiatives. He currently leads efforts to focus on the importance of software as a first-class scholarly artifact and explores all aspects of how libraries might become critical partners in developing strategies and techniques for the future of research traceability and reproducibility through software, data, computational narratives and other scientific digital artifacts. 

Event Category:

Novel Parameterizations of sub-grid scale aerosol-cloud processes for climate models

Location: 
CG Auditorium
Speaker: 
Daniel Rothenberg

Aerosol-cloud interactions are mediated in global climate models with the help of aerosol activation parameterizations. These parameterizations translate how changes in the background, ambient aerosol simulated by the model influence the nucleation of cloud droplets - specifically the number concentration of how many are nucleated. Altering aerosol emissions in the model thus influences cloud droplet number concentrations, giving rise to an “aerosol indirect effect” on climate as the simulated clouds exhibit different radiative properties.

Speaker Description: 

Dr. Rothenberg is the Chief Scientist at ClimaCell, a weather technology company which leverages novel atmospheric measurements to better understand the weather as it impacts people and businesses. Previously, he completed his doctoral and post-doctoral studies at MIT where his dissertation focused on better understanding how aerosols influence climate. In this work he leveraged data-driven modeling and machine learning to build process emulators which could be embedded in climate models to more faithfully represent the physics underlying aerosol-cloud interactions.

Event Category:

CodeReef: an open portal for cross-platform MLOps and reproducible benchmarking

Location: 
CG Auditorium
Speaker: 
Hervé Guillou

We present CodeReef - an open platform to share all the compo-
nents necessary to enable cross-platform MLOps, i.e. automating
the deployment of ML models across diverse systems in the most
efficient way. We also introduce the CodeReef solution - a way to
package and share models as portable, reproducible and customiz-
able archive files. Such ML packs include JSON meta descriptions,
Python APIs, CLI actions and portable workflows necessary to au-
tomatically build, benchmark, test and customize models across

Speaker Description: 

Machine learning Engineer at CodeReef since september 2019.Previously, I have been working for 9 years at CEA as a data scientist and machine learning engineer. My job was to implement the latest ML algos to optimize the consumption of renewable energy on GRID. 

Event Category:

Exploring Scientific Datasets using Collaborative, Immersive Visualization

Location: 
CG Auditorium
Speaker: 
Nicholas Brunhart-Lupo

We discuss the value of collaborative, immersive visualization for the exploration of scientific datasets and review techniques and tools that have been developed and are deployed at the National Renewable Energy Laboratory (NREL). We believe that collaborative visualizations that link statistical interfaces on laptops, high-performance computing (HPC), virtual reality (VR) headsets, and cave automatic virtual environments (CAVEs) enable scientific workflows that further rapid exploration of large, high-dimensional datasets by teams of analysts.

Speaker Description: 

Nicholas Brunhart-Lupo is a visualization scientist at the National Renewable Energy Laboratory. He received a PhD degree in computer science from the Colorado School of Mines and a master's degree in computer science from the University of Queensland, Brisbane. His research interests include 3D vector field topology, scientific visualization techniques in immersive spaces, and graphics programming technologies.

Event Category:

ecProf -- a non-intrusive MPI+OpenMP profiler

Location: 
CG Auditorium
Speaker: 
Sami Saarinen

The ecProf profiler is capable for performance analysis (including PAPI counters), I/O and memory tracing and has been long time coming from ECMWF. This profiler can analyze any Linux dynamic executables (including Linux commands) without any need to add explicit calls to the application program, or even to have the application source code available. It knows about MPI (if present), and OpenMP (if present) and can also trace presence of any auxiliary pthreads.

Speaker Description: 

Senior Analyst and Team Leader in HPC Section's application software support

Event Category:

SOFTWARE ENGINEERING METHODS USED FOR THE PARALLELIO (PIO) C/FORTRAN LIBRARIES

Location: 
CG Auditorium
Speaker: 
Ed Hartnett

The PIO C and Fortran libraries allow for high-performance I/O on HPC systems. These libraries are developed using software engineering techniques such as branch development, pull-requests, automated testing, continuous integration, portable releases which adapt to user installation conditions, and full documentation of code for users and developers. This paper details the use of software engineering techniques on the PIO software project.

Speaker Description: 

I am working on HPC I/O. I am a developer on the PIO library, and also netcdf. I have also lead software engineering teams in industry and as ground data processing for several NASA Earth observation and planetary missions.

Event Category:

Cryohub: Jupyterhub Deployments for the Cryosphere

Location: 
CG Auditorium
Speaker: 
Luis Lopez

Workflows to analyze large scientific datasets have historically followed a search-subset-download pattern. This traditional approach has a lot of disadvantages, from research being compartmentalized and not easily reproducible to a lack of scalability and portability. Jupyterhub is a relatively new platform in the data science world that encourages reproducibility and lowers the entry barrier to non-expert users by eliminating individual prerequisites and machine dependencies needed to share data analysis workflows .

Speaker Description: 

Software Engineer with a master's degree in computer science from the University of Colorado and interests in distributed systems, cloud computing and machine learning.

Event Category:

Pages

Subscribe to conference-talk