AMSC 663-664 Catalog Descriptions

AMSC 664, Advanced Scientific Computing II (3 credits):
In the sequence AMSC 663-664 students work on a year-long individual project to develop software for a scientific task in a high perfomance computing environment. Lectures will be given on code development and validation, parallel algorithms for partial differential equations, nonlinear systems, and optimization.

Details and Syllabi

Advanced Scientific Computing I (ASC I)

In this course topics covered in SC I, SC II, and COPSC are applied in the context of a year-long project to develop software designed to carry out a computational scientific task in a high performance computing environment. In this term the project includes a written and oral proposal within one month, and a written and oral progress report at the end of the term.

Prerequisites and Co-requisite

• SC I and SC II are prerequisites
• COPSC is a co-requisite
• permission of an instructor

Syllabus

• Project Design Proposal
  • By the end of one month each student must find a faculty advisor and identify a suitable project that includes a deliverable suite of software that is designed to carry out a computational scientific task, propose appropriate algorithms, languages, and platforms for the development of this software, write a short project proposal that also includes a scientific justification, and present the proposal orally.
• Available Computational Environments
  • While the students are formulating their proposals, they will taught specific capabilities of the various high performance computational and graphical platforms available to them. This includes the attendant networks and their communication tools.
• Code Development
  • modularity, portability, memory management
  • post-processing, restarting, and writing to databases
  • interactivity
  • scientific visualisation
  • documentation and version management tools
  • debugging and profiling tools
• Implementation of Parallel Algorithms
  • use of distributed memory and local caches
  • masking communication costs, load balancing, granularity
  • parallel numerical linear algebra for full and sparse matrices
  • basic parallel algorithms for PDE e.g., simple domain decompositions
• Project Progress Report
  • Each student must give a written and oral report on the state of his or her project, explain how the software has been developed and tested, give his or her current vision of the finished product, and detail how that vision has evolved over the course of the project.

Advanced Scientific Computing II (ASC II)

This course is a continuation of ASC I. This course focuses on issues of code development and validation relevant to large scientific computing projects on high performance computing systems. In this term the project includes a seminar lecture and a written final report.

Prerequisites

• ASC I

Syllabus

• Code Development and Validation
  • round-off sensitivity, performance analysis, tuning to platforms
  • symmetry and stability studies
  • analytical, experimental, and computational benchmark comparisons
  • convergence studies
  • parametric sensitivity studies
  • investigation of asymptotic regimes
• Advanced Parallel Computing
  • parallel algorithms for PDE, e.g., adpative mesh refinement, multigrid, independent time steps, domain decomposition, operator splitting
  • parallel algorithms for nonlinear systems and optimization
• Computational Science Seminars
  • Local and visiting computational scientists representing a broad selection of scientific disciplines will explain the computational aspects of their work. These aspects should be reflected in the choice of topics covered above. Each student is expected to listen to these seminars critically.
• Project Final Report
  • Each student must give a one hour seminar and a complete written report on his or her project, explain how the software has been developed, tested and validated, document the software and how it is used, show results of its use, detail how his or her vision of the finished product has evolved over the course of the project.