DOC PREVIEW
UW-Madison ME 964 - Syllabus

This preview shows page 1-2 out of 5 pages.

Save
View full document
View full document
Premium Document
Do you want full access? Go Premium and unlock all 5 pages.
Access to all documents
Download any document
Ad free experience
View full document
Premium Document
Do you want full access? Go Premium and unlock all 5 pages.
Access to all documents
Download any document
Ad free experience
Premium Document
Do you want full access? Go Premium and unlock all 5 pages.
Access to all documents
Download any document
Ad free experience

Unformatted text preview:

Version: 5/8/2012 Page 1 of 5 ME 964 – Spring 2012 High Performance Computing for Engineering Applications Time: 9:30 – 10:45 Tu&Th Location: 1163ME Instructor: Dan Negrut Office: 2035ME Phone: 608 890 0914 E-Mail: [email protected] Course Page: http://sbel.wisc.edu/Courses/ME964/2012/index.htm Grades Page: http://learnuw.wisc.edu Video Streaming: http://mediasite.engr.wisc.edu/Mediasite/Catalog/pages/catalog.aspx?catalogId=31c0b7c4-3a0f-410b-bacf-0c238380112f&folderId=96ee9eab-32a4-4321-8b45-6eae85c267ef&rootDynamicFolderId=e5b4a945-c68f-45b2-9eb7-b2512f5122cd Forum: http://sbel.wisc.edu/Forum/viewforum.php?f=3 T.A.: Toby Heyn ([email protected]) Office Hours: Monday, 2 – 4 PM Wednesday, 2 – 4 PM Other times by appointment (please call or email to arrange) Prerequisites: C Programming Recommended Texts (see also lecture handout for other recommendations): Jason Sanders and Edward Kandrot: CUDA by Example: An Introduction to General-Purpose GPU Programming, Addison-Wesley Professional, 2010 (on reserve at Wendt Library) GPU Programming Guide: http://developer.download.nvidia.com/compute/cuda/4_0/toolkit/docs/CUDA_C_Programming_Guide.pdf, download from NVIDIA website, David B. Kirk and Wen-mei W. Hwu: Programming Massively Parallel Processors: A Hands-on Approach, Morgan Kaufmann, 2010 (on reserve at Wendt Library) H. Nguyen (ed.), GPU Gems 3, Addison Wesley, 2007 (on reserve at Wendt Library) Peter Pacheco: An Introduction to Parallel Programming, Morgan Kaufmann, 2011 T. Mattson, et al.: Patterns for Parallel Programming, Addison Wesley, 2005 Michael J. Quinn: Parallel Programming in C with MPI and OpenMP, McGraw Hill, 2003 Course Objectives: The course is meant to (a) provide an overview of existing High-Performance Computing (HPC) software and hardware, (b) present basic software design patterns for high performance parallel computing, (c) introduce CUDA for parallel computing on the Graphics Processing Unit (GPU), (d) introduce the Message Passing Interface (MPI) standard and its MPICH2 implementation for leveraging parallelism on a CPU cluster, and (e) introduce the OpenMP solution to enabling parallelism across multiple CPU cores. The approach is hands-on, the students are expected to use the lecture information, a series of assignments and a final project to emerge at the end of the class with parallel programing knowledge that can be immediately applied to their research projects.Version: 5/8/2012 Page 2 of 5 Hardware Used: The course is designed to leverage Euler, a cluster with about 200 CPU cores and 56 GPU cards. Each student will receive an individual account on Euler that will be used for GPU computing, MPI-enabled parallel computing and possibly OpenMP multi-core computing. Euler runs Scientific Linux. Course Workload: The course will have one midterm exam, one midterm project, and one final project. All projects will be either individual or two-member team projects. Course Structure: For this class, the Microsoft PowerPoint notes used in class will be posted online at http://sbel.wisc.edu/Courses/ME964/2012/index.htm. Grades in ME964 will be based on your performance on homework, midterm exam, one midterm project, one final project, and course participation. All homework and exam scores will be maintained on the Learn@UW course website. This will allow you to monitor your performance and see aggregate scores for the rest of the class, which can give you a continuous idea of your performance in relation to the rest of the class. Should you have questions about your score, please contact the instruction. Policies regarding grading and turning in your homework: 1. Score-related questions about homeworks, midterm exam, and midterm project must be raised prior to the next class period after receiving the score. 2. If homework that you turned in appears not to be graded (missing) on the Learn@UW course website please point that out to me within one week after the return of the corresponding set of graded homeworks. It is a good practice to save your homework so that I will be able to update the grade to give you full credit for your work. 3. The homework with the lowest score will be dropped when computing the final homework average Final Grade Policy: The final grade will be computed using the following weights: Homework = 40% Midterm Exam = 10% Midterm Project = 20% Final Project = 25% Course Participation = 5% Homework: Assignments will be handed out roughly on a weekly basis for the first part of the class. The homework solution will be returned using the Lear@UW drop-box system. Homework solutions should be neat, well organized, and well commented. Your score for each assignment will be between 0-100. No late homework will be accepted. The homework with the lowest score will be dropped when computing final score. Midterm Exam: There will be one “closed-books” midterm exam that will cover material introduced in the first part of the course. You can bring along annotated copies of the documents that you have been asked to read (reading assignments). There will be no need for a computer for this test. The best way to prepare for exams is to participate in class, learn the fundamental concepts, and work on the assignments diligently. The exam will be scored on a scale of 1 to 100. Note that there will be no final exam. Midterm Project: You will have to select by March 8 the topic of your Midterm Project. The Midterm Project software and technical report is due on April 12 at 11:59 PM. All students are expected to work on the project individually or in two-person teams. The topic of this project will be the solution ofVersion: 5/8/2012 Page 3 of 5 either dense banded or fully banded linear systems in CUDA. You will have the option to choose one or the other. Your project should include your CUDA code, profiling results, weak scaling analysis, and a comparison with either a sequential or a parallel state-of-the-art-solution available commercially or as open source. Final Project: The projects will be individual or generated by two-student teams. The topic of the project could be related to your research. Alternatively, you can continue to work on and augment your


View Full Document

UW-Madison ME 964 - Syllabus

Documents in this Course
Load more
Download Syllabus
Our administrator received your request to download this document. We will send you the file to your email shortly.
Loading Unlocking...
Login

Join to view Syllabus and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Syllabus 2 2 and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?