Lecture (After Call) FZJ-2018-04961

http://join2-wiki.gsi.de/foswiki/pub/Main/Artwork/join2_logo100x88.png
High Performance and Cloud Computing for Remote Sensing Data

 ;  ;

2018

Lecture at IEEE International Geoscience and Remote Sensing Symposium (IGARSS) (Valencia, Spain), 21 Jul 2018 - 21 Jul 20182018-07-212018-07-21

Please use a persistent id in citations:

Abstract: The development of the latest-generation sensors mounted on board of Earth observation platforms has led to a necessary re-definition of the challenges within the entire lifecycle of remote sensing data. The acquisition, processing and application phases face problems, which are well described by the Vs big data definitions: (1) Volume - the increasing scale of archived data (i.e., hundreds of Terabytes per day) raises not only data storage but also massive analysis issues, (2) Variety - the data are delivered by sensors acting over different spatial, spectral, radiometric and temporal resolutions, (3) Velocity - the data processing and analysis must confront the rapidly growing rate of data generation, (4) Veracity - the massive amount of sensed data coming in at high speed is associated with uncertainty and accuracy measurement and (5) Value - the acquired data are not straightforward to interpret and they require a powerful yet highly accurate processing scheme in order to extract reliable and valuable information. Traditional serial methods (i.e., desktop approaches, such as MATLAB, R, SAS, etc.) present several limitations and they become ineffective when considering these challenges. Despite modern desktop computers and laptops becoming increasingly multi core and performing better, they preserve limitations in terms of memory and core availability. Trends in parallel architectures like many-core systems (e.g. GPUs) are in continuous expansion to satisfy the growing demands of domain-specific applications for handling computationally intensive problems. In the context of large scale remote sensing applications, High Performance Computing (HPC) and Cloud Computing have the chance of overcoming the limitations of serial algorithms. Parallel architectures such as clusters, grids, or clouds provide tremendous computation capacity and outstanding scalability underpinned by strong and stable standards used for decades, e.g. message passing interface (MPI). The tutorial aims at providing a complete overview for an audience that is not familiar with these topics. The tutorial will follow a twofold approach: selected background lectures (morning session) followed by practical hands-on exercises (afternoon session) in order to enable you to perform your own research. The tutorial will discuss the fundamentals of supercomputing as well as cloud computing, and how we can take advantage of such systems to solve remote sensing problems that require fast and highly scalable solutions.


Contributing Institute(s):
  1. Jülich Supercomputing Center (JSC)
Research Program(s):
  1. 512 - Data-Intensive Science and Federated Computing (POF3-512) (POF3-512)
  2. DEEP-EST - DEEP - Extreme Scale Technologies (754304) (754304)
  3. PhD no Grant - Doktorand ohne besondere Förderung (PHD-NO-GRANT-20170405) (PHD-NO-GRANT-20170405)

Appears in the scientific report 2018
Database coverage:
OpenAccess
Click to display QR Code for this record

The record appears in these collections:
Document types > Presentations > Lectures
Workflow collections > Public records
Institute Collections > JSC
Publications database
Open Access

 Record created 2018-08-16, last modified 2021-01-29


OpenAccess:
JSC-Intro - Download fulltext PDF Download fulltext PDF (PDFA)
HPC for Big Remote Sensing Data Analytics (Support Vector Machines) - Download fulltext PDF Download fulltext PDF (PDFA)
HPC for Big Remote Sensing Data Analytics (Convolutional Neural Networks) - Download fulltext PDF Download fulltext PDF (PDFA)
Parallel Programming (MPI) and Batch Usage (SLURM) - Download fulltext PDF Download fulltext PDF (PDFA)
External link:
Download fulltextFulltext by OpenAccess repository
Rate this document:

Rate this document:
1
2
3
 
(Not yet reviewed)