CSCE 569 Parallel Computing, Fall 2015

Course-flier.pdf

Department of Computer Science and Engineering
University of South Carolina

Course Meeting Time & Location MWF 1:10PM- 2:00PM
Classroom: 2A21 Swearinger Engineering Center

Textbooks

Introduction to parallel programming By:Peter S. Pacheco Publisher: Morgan Kaufmann
Language: English
ISBN-10: 0123742609
ISBN-13: 978-0123742605

Instructor: Dr. Jianjun Hu
Machine Learning and Evolution Group (MLEG)
Email: jianjunh AT cse.sc.edu
Office: 3A66 Swearinger Engineering Center
Office Hours: MW 2:30PM-3:30PM or by Appointment.

Course Description This course will cover the techniques and topics that are widely used in real-world parallel computing. This is about learning how to run your programs on hundreds of computers or thousands of CPUs to solve real-world large problems. This is about how to handle large-scale data processing as those used in Google. Students in science and engineering will all benefit from this course as scientific computing has become one of the main ways for discovery and invention.

Course Highlights Many interesting applications, Hands-on projects, exposure to latest techniques. Topics include but not limited to:

  • Linux cluster computing
  • PBS systems
  • Parallel programming using MPI, OpenMP, pthreads

Prerequisite You should be able to program using high-level language C/C++ or java.

If you have problem to enroll, pls contact/call CSE secretary

Randi 803 777-7849 (phone) Swearingen Bldg., Room 3A01L

Supercomputers at USC College of Engineering and Computing(from HPC)

  • Nick Linux clusters with 291 CPUs
  • Optimus Linux clusters with 256 CPUs
  • Zia Share memory computer with 128 CPUs and 256 Shared memory
  • Nataku, Jaws2 (Chemistry Department) for fuel cell simulation and etc.

Blue-gene SuperComputer with 212992 CPUs

Department of Computer Science and Engineering
College of Engineering and Computing
University of South Carolina
301 Main Street, Columbia, SC, 29201

http://mycounter.tinycounter.com/index.php?user=gespim