Document Type
Poster
Conference Title
Oklahoma Research Day 2017
City and State
Enid, OK
Conference Date
March 3, 2017
Publication Date
3-3-2017
Abstract
The National Aeronautics and Space Administration (NASA) produces nearly two gigabytes of data per second. NASA researchers leverage millions of dollars of computing hardware to analyze this data. NASA shares this data with the rest of the world. Advances in computer technology have provided modern desktop computers more powerful than the fastest supercomputers in the world from two and three decades ago. This provides many possibilities for greater use of NASA data. A lack of education materials for undergraduate research in high performance computing limits these possibilities. This research addresses this need by presenting the methodologies used to translate the NASA MAPSS software system from supercomputers and software engineers to desktops and undergrads. Undergraduate student researchers studied the MAPSS software system, created for the NASA Goddard Space Flight Center supercomputers, to conduct a validation study of NASA Earth Atmosphere aerosol data. Undergrads rewrote parts of the software allowing it to run on an Intel \1 processor running a Linux system. The students completed translation of four of the seven satellite sensors, and developed automation software allowing MAPSS to be portable between individual computers. The students provided documentation of this process allowing future students to complete the translation of the remaining sensor systems and the validation study. This should provide greater use of the data that streams from NASA every day.
Recommended Citation
Sleeper, Charles and Evert, Jeremy, "Porting a parallel program from the NASA Center for Climate Simulation (NCCS) Discoverer supercomputer to desktops for validation of the Multi-sensor Aerosol Products Sampling System (MAPSS)" (2017). Student Research. 5.
https://dc.swosu.edu/cpgs_edsbt_bcs_student/5