High Performance Computing: An Introduction

Prerequisites

The course aims to give an introductory overview of High Performance Computing (HPC) in general, and of the facilities of the High Performance Computing Service (HPCS) available at the University of Cambridge.

Practical examples of using the HPCS clusters will be used throughout, although it is hoped that much of the content will have applicability to systems elsewhere.

The training room is located on the first floor and there is currently no wheelchair or level access available to this level.

Please note that if you are not eligible for a University of Cambridge Raven account you will need to book or register your interest by linking here.

• Novice users of HPC and anyone who expects to need to use HPC systems at some stage in their research
• Graduate students, Postdocs and Staff members from the University of Cambridge, Affiliated Institutions and other external Institutions or individuals
• Please be aware that these courses are only free for registered University of Cambridge students. All other participants will be charged a registration fee in some form. Registration fees and further details regarding the charging policy are available here.
• Further details regarding eligibility criteria are available here

• Attendance at a Unix introductory course (or equivalent knowledge) such as
  • Introduction to Unix Shell (face to face)
  • Unix: Introduction to the Command Line Interface (Self-paced)
• It is desirable to have shell scripting experience equivalent to Unix: Simple Shell Scripting for Scientists

Number of sessions: 1

Basic concepts:

• serial, parallel and high throughput workload
• a quick look at the modern computer (vector/multithread/multicore/multisocket CPUs; nonuniform memory)
• interconnecting nodes (ethernet, Infiniband, proprietary)
• how to put it together (shared memory, distributed memory, ccNUMA, single system image)
• coprocessors (GPUs and similar things)
• cluster storage
• job scheduling

High Performance Computing Service:

• Peta 4 - an example of an infiniband CPU cluster
• Wilkes 2- an example of a dual-rail infiniband GPU cluster
• Service Levels (free and non-free usage)
• Help and further information
• Connecting
• SSH login
• File transfer
• Remote desktop
• Tunnelling

Security:

• Client side
• Server side

User environment:

• Compilers
• Environment modules
• Filesystems

Software:

• Free
• Proprietary

Job submission:

• Batch scheduler (SLURM)
• How to submit (HTC, MPI, OpenMP, hybrid)
• Array jobs
• Interactive jobs
• Checkpoint/restart

Presentations and practicals

• Free for registered University of Cambridge students
• £ 50/day for all University of Cambridge staff, including postdocs, temporary visitors (students and researchers) and participants from Affiliated Institutions. Please note that these charges are recovered by us at the Institutional level
• It remains the participant's responsibility to acquire prior approval from the relevant group leader, line manager or budget holder to attend the course. It is requested that people booking only do so with the agreement of the relevant party as costs will be charged back to your Lab Head or Group Supervisor.
• £ 50/day for all other academic participants from external Institutions and charitable organizations. These charges must be paid at registration
• £ 100/day for all Industry participants. These charges must be paid at registration
• Further details regarding the charging policy are available here

Programming expertise is not required (the focus will be on how to run, rather than on what to run)

1

A number of times per year

• Unix: Introduction to the Command Line Interface (Self-paced)   [Provided by University Information Services - Digital Literacy Skills]
• Unix: Simple Shell Scripting for Scientists   [Provided by University Information Services - Digital Literacy Skills]