ACENET Parallel Computing School

June 2, 4, 9, 11, 16, 18, 23, 25, 30, July 2, 7, 9, 2020

2:00PM - 4:00PM

Instructors: Ross Dickson, Mat Larade, Serguei Vassiliev

Helpers: Grace Fishbein, TBD

This site contains the general information for the 2020 Parallel Computing School, put on by ACENET.

General Information

The ACENET Parallel Computing School seeks to educate researchers in those tools and techniques used in high-performance computing and scientific computation. Topics will include general parallel computing, OpenMP, GPGPU, and Message Passing Interface (MPI).

Who: The course is aimed at researchers, both academic and industrial. The background expected is typically that of a graduate student, although both advanced undergraduates and those who have finished (or never been to) graduate school may expect to benefit. Participants attending the school must have familiarity with the Unix command line, and some level of programming experience. Participating in the ACENET Basics Series would be a great foundation for the school.

Requirements:

Setup

Zoom

This online training will be conducted using Zoom video conferencing. Please download and install Zoom Client for Meetings.

SSH client

You must have an SSH client installed on your computer, and be able to connect to a Compute Canada cluster with it.

Contact: Please email ross.dickson@ace-net.ca for more information.


Schedule (under construction)

Tuesday, June 2

14:00-16:00 Atlantic Generalized parallel Computing

Thursday, June 4

14:00-16:00 Atlantic Working at scale

Tuesday, June 9

14:00-16:00 Atlantic Dask 1

Thursday, June 11

14:00-16:00 Atlantic Dask 2

Tuesday, June 16

14:00-16:00 Atlantic Open MP Shared Memory

Thursday, June 18

14:00-16:00 Atlantic Open MP Shared Threads

Tuesday, June 23

14:00-16:00 Atlantic Machine Learning Intro/Classification

Thursday, June 25

14:00-16:00 Atlantic Machine Learning Regression/Neural Networks

Tuesday, June 30

14:00-16:00 Atlantic GPGPU 1

Thursday, July 2

14:00-16:00 Atlantic GPGPU 2

Tuesday, July 7

14:00-16:00 Atlantic MPI 1

Thursday, July 9

14:00-16:00 Atlantic MPI 2

Syllabus

General parallel computing and working at scale

  • performance measurement, and factors affecting performance
  • I/O and data movement
  • schedulers and “embarrassingly parallel” jobs
  • analyzing performance using a profiler
  • Reference...

OpenMP, shared memory, and threads