ITaP staff presenting at advanced research computing conference
When the Practice and Experience in Advanced Research Computing (PEARC) conference convenes in New Orleans next week, Purdue will be all over the agenda.
ITaP Research Computing staff will give presentations and lead interactive workshops about a variety of topics, including a novel new allocation system for ITaP’s Data Depot research data storage service, geospatial data visualization tools and web-based science gateways. They also will talk about work done with the National Science Foundation’s Extreme Science and Engineering Discovery Environment (XSEDE) and Purdue’s efforts to increase the representation of women in high-performance computing.
Dan Dietz, an ITaP Research Computing senior scientific applications analyst, will highlight the implementation and architecture of the new Data Depot storage allocation system, which can fulfill purchase orders instantly with the click of a button. The old system required submitting a ticket to storage administrators and waiting for them to make the allocation. The new system is an example of ITaP Research Computing’s efforts to shorten researchers’ time to science.
The Geospatial Data Analysis Building Blocks (GABBs) team of Carol Song, Rajesh Kalyanam, Derrick Kearney, Lan Zhao, Carolyn Ellis, Rob Campbell and Larry Biehl will present their efforts to cloud-enable a science gateway that supports geospatial data management, visualization, processing and publication. The end product allows anyone to use GABBs components on their own web-based hub powered by HUBzero, Purdue’s open-source platform for scientific collaboration.
Along with Michael Zentner, ITaP senior research scientist, the GABBs team will also lead an interactive tutorial about the use of geospatial data visualization tools within HUBzero. The tutorial will include a hands-on demonstration of tools created by the GABBs project for geospatial visual analytics, and will illustrate the added value of working from the HUBzero platform. By the end of the tutorial attendees will be able to create and share Jupyter notebooks and applications built in RStudio on a HUBzero hub.
In addition, HUBzero senior software engineer Steve Clark will speak on a panel about best practices for science gateway job management.
ITaP Research Computing program manager Marisa Brazil serves as champion coordinator for XSEDE’s Campus Champions program, which connects and supports 340 individuals at more than 200 institutions nationwide who help researchers make use of XSEDE resources. Campus Champions activities at PEARC include an evening networking event, a session for champions to get together and discuss topics of shared interest, and meetings for the leadership team and regional leaders.
An important initiative of Campus Champions is the Campus Champions Fellows program, which pairs champions with high-end application support staff in XSEDE’s Extended Collaborative Support Services (ECSS). ITaP Research Computing data visualization specialist Tsai-wei Wu is one of five fellows for 2016-2017 and has spent the last year working on a project modeling two-dimensional semiconductors with her mentor Sudhakar Pamidighantam of Indiana University. She will present her work at PEARC.
Brazil and Gladys Andino, an ITaP Research Computing senior scientific applications analyst, along with Preston Smith, director of research services and support for ITaP, and Michael Gribskov, a professor of biological sciences, will present a poster about Purdue’s efforts to increase the representation of women in the field of high-performance computing. Their poster describes the HPC expertise of each of the female computational domain scientists working for ITaP Research Computing, as well the launch of Purdue’s Women in HPC networking group, which connects student, faculty and staff women with an interest in high-performance computing. The group, started by Andino last fall, now has more than 50 members.
With support from Women in HPC’s student travel scholarship, Song-hui Ryu, a graduate student in computer and information technology, will present her poster titled “Comparison of Machine Learning Algorithms for Botnet Detection” at PEARC.