High Performance Computing Summer Boot Camp

High Performance Computing (HPC) Summer Boot Camp – OpenACC, OpenMP, OpenMPI

In conjunction with the Pittsburgh Supercomputer Center (PSC), Tufts is offering a week long summer boot camp bringing together many of the technologies in High Performance Compute (HPC) programming. GPU accelerators (OpenACC), single node scaling (OpenMP) and cluster wide scaling (OpenMPI) will be covered along with parallel development, debugging tools. The week will culminate with a discussion of hybrid computing which brings everything together as well as post boot camp competition. Seats will go fast, so sign up early and join the Research Technology (RT) staff for an excellent week of free training. These workshops are sponsored by the National Science Foundation (NSF) funded eXtreme Science and Engineering Discovery Environment (XSEDE).

Sign-up at: XSEDE HPC Summer Boot Camp

Date: Four Days, Monday – Thursday, June 4-7, 2018. 11:00 a.m.-5:00 p.m.

Location: Collaborative Learning and Innovation Complex, Room 316

Address574 Boston Ave Medford, MA 02155

Audience: All faculty, student and staff

Agenda:

  • Monday, June 4
  • 11:00 Welcome
  • 11:15 Computing Environment
  • 11:45 Intro to Parallel Computing
  • 12:30 Intro to OpenMP
  • 1:30 Lunch break
  • 2:30 Exercise 1 (zip folder of all exercises)
  • 3:15 More OpenMP
  • 4:30 Exercise 2
  • 5:00 Adjourn
  • Tuesday, June 5
  • 11:00 Intro to OpenACC
  • 12:00 Exercise 1
  • 12:30 Introduction to OpenACC (cont.)
  • 1:00 Lunch break
  • 2:00 Exercise 2
  • 2:45 Introduction to OpenACC (cont.)
  • 3:00 Using OpenACC with CUDA Libraries
  • 3:30 Advanced OpenACC
  • 4:00 OpenMP 4.0 Sneak Peek
  • 5:00 Adjourn
  • Wednesday, June 6
  • 11:00 Introduction to MPI
  • 1:00 Lunch break
  • 2:00 Intro exercises
  • 3:10 Intro exercises review
  • 3:15 Scalable Programming: Laplace code
  • 3:45 Laplace Exercise
  • 5:00 Adjourn
  • Thursday, June 7
  • 11:00 Laplace Exercise (Cont)
  • 12:30 Laplace Solution
  • 1:00 Lunch break
  • 2:00 Advanced MPI
  • 3:00 Outro to Parallel Computing
  • 4:00 Parallel Tools
  • 4:20 Hybrid Computing
  • 4:40 Hybrid Competition
  • 5:00 Adjourn