Thesis
Communication-avoiding ray tracing for exascale computing
Washington State University
Master of Science (MS), Washington State University
2016
Handle:
https://hdl.handle.net/2376/103717
Abstract
Exascale computers, defined as being capable of performing at least one exaFLOPS (1018 floating point operations per second) are anticipated to emerge in the next several years. Reaching this scale of computation will require hardware and software changes for high-performance computing (HPC). Applications will need to adapt as the architecture of supercomputers change. Some studies suggest communication-avoiding algorithms might be the most performant design for future systems. This has created an interest in the development of scalable visualization algorithms and techniques. This thesis explores the impact exascale hardware will have on programming models and application design. We look at one specific programming model, Concurrent Collections (CnC), and explore how we can use it coupled with Intel's ray tracing engine, Embree, to build a scalable ray tracing system with an emphasis on communication avoidance and extension for exascale.
Metrics
11 File views/ downloads
16 Record Views
Details
- Title
- Communication-avoiding ray tracing for exascale computing
- Creators
- Ellen Alice Porter
- Contributors
- Robert R. Lewis (Degree Supervisor)
- Awarding Institution
- Washington State University
- Academic Unit
- Electrical Engineering and Computer Science, School of
- Theses and Dissertations
- Master of Science (MS), Washington State University
- Publisher
- Washington State University; [Pullman, Washington] :
- Identifiers
- 99900525061401842
- Language
- English
- Resource Type
- Thesis