High Performance Computing Microcredential – Digital Nova Scotia – Leading Digital Industry

September 15 – October 27, 2026

6-week program, 2 X 2-hour sessions/week  |  24 hours total training

Valued at $1,225.00 + HST
Eligible DNS members pay $310 + HST

Hosted by acenet

More CoursesApply Now

This course combines virtual classroom teaching and an independent study project. Participants will learn the fundamentals of high-performance computing and apply those skills on ACENET’s computing cluster.

High-Performance Computing (HPC) refers to the use of supercomputers with thousands of processing units to solve tasks that are too large and/or complex for a personal laptop or desktop. Breaking a large problem into tens, hundreds, or even thousands of smaller problems to be solved in parallel reduces processing time to a fraction of what it would be on a desktop or workstation.

HPC is widely used to scale up code and optimize performance in large data mining, training of AI models, engineering and scientific simulations, remote sensing, land and ocean mapping, and many other fields.

What will you learn? (More information in the FAQ section)
    • How to securely access and work within high-performance computing (HPC) environments.
    • How to run and manage large-scale computing workloads using batch scheduling tools.
    • How to profile and optimize code performance for faster, more efficient computation.
    • How to build parallel programs with OpenMP and CUDA.
    • How to deploy and manage HPC clusters on cloud platforms.
    • How to leverage ACENET’s supercomputing infrastructure and connect with Atlantic Canada’s tech ecosystem.
Who’s right for this course?

This intermediate level course is designed for mid-career IT professionals. This course will target individuals with the following prerequisites:

  • Proficiency with programming and IT development
  • Proficiency with Linux
  • Familiarity with Github (asynchronous learning material will provided prior to course commencement)
  • Basic familiarity with the C programming language (either a single university-level course, or self-taught are sufficient).

About the Instructor

Ross Dickson

Sarah Clarke

Serguei Vassiliev

Funding

To apply your organization must

Funding

Digital Nova Scotia will facilitate the funding process based on the information provided in the application below. Employees must be receiving full-time compensation while in training to be eligible for WIPSI funding. This may include employees who are full-time seasonal.

Apply Now

Please complete the following form to confirm your eligibility for funding, including uploading the Business Information and Registration forms. Please give us at least 3 – 5 business days to respond to your email with your eligibility.

Have Questions?

Please contact Emily van den Hoogen at emily@digitalnovascotia.com

Course FAQ

[ click on a question to view the answer ]

What will the training cover?

Course Curriculum Overview:

  • Session 1: Description of HPC; accessing remote resources; batch scheduling (Learning Objectives 1-4)
  • Session 2: Performance profiling I (Learning Objective 5)
  • Session 3: Mathematical models of strong scaling and weak scaling (Learning Objective 6)
  • Session 4: OpenMP programming I (Learning Objective 7)
  • Session 5: OpenMP programming II (Learning Objective 7 continued)
  • Session 6: GPU programming I (Learning Objectives 9, 10)
  • Session 7: GPU programming II (Learning Objective 10 continued)
  • Session 8: GPU programming III (Learning Objective 11)
  • Session 9: HPC on cloud services (Learning Objective 12)
  • Session 10: Performance profiling II (Learning Objective 5)
  • Session 11: Independent study project presentations
  • Session 12: Independent study project presentations

Program Outcomes

By the end of this course, participants will understand how:

  1. an HPC cluster is different from personal computers and other “cloud computing” resources.
  2. to use SSH keys to access HPC resources securely.
  3. to use a batch scheduler to share HPC resources.
  4. hardware affects parallelism and how to make programming decisions.
  5. to profile a program to measure speed and identify hot spots or bottlenecks.
  6. to apply the concepts of strong and weak scaling, and the mathematical models describing them.
  7. to make an existing C language program multi-threaded using OpenMP.
  8. to identify data dependencies, race conditions, and other hazards of parallel programming.
  9. a GPU differs from a CPU.
  10. to use CUDA to program a GPU.
  11. to port an existing CPU program to use a GPU.
  12. to create an HPC cluster on a commercial cloud service (AWS, Google Cloud, Azure).