 | Use this page to maintain syllabus information, learning objectives, required materials, and technical requirements for the course. |
| CNIT 56100 - Advanced Parallel Data Systems |
|---|
|
Associated Term:
Spring 2013
Learning Outcomes: 1. Understand the impact of performance, cost, reliability, and usability on the design and deployment of high performance computing systems based on mixture of commodity and special purpose components and software. 2. Demonstrate skill in finding a balance among these factors by designing, building, benchmarking, and optimizing a small commodity-based cluster computer based on the Linux operating system and other open source software packages. 3. Demonstrate ability to analyze problems inherent in cluster computing and to design new solutions to those problems based on the development and integration of new technologies, such as GPUs, parallel data systems, and workflow systems. 4. Demonstrate understanding and knowledge of the core concepts of high performance computing, which include: The continuum of high performance computing architectures, and the appropriateness of each architecture type to problem solving for a wide variety of applications; The effects of communications architecture and performance characteristics on application performance; The effects of component reliability and operating systems on Petascale computing; The use of commodity components in high performance computing, and the trends and forces that motivate their use; The design of high performance computing systems to meet specific application needs and resource constraints; tools and techniques of cluster and high performance computing, such as scheduler configuration and use, benchmarking tools and techniques, queuing and reliability models. Required Materials: Technical Requirements: |
| Return to Previous | New Search |
 |