Skip to content

Latest commit

 

History

History

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 

README.md

Faster Computing Part 2

Today we will make use of a high performance computing (HPC) cluster. HPC clusters provide many more servers (nodes) on which to run things in parallel. Moreover, each node is also faster than your personal computer. Thus, even if you're not running code in parallel, it is often faster to run things on a single node of the cluster than on your computer. (Plus it frees you up to do other things with your computer!)

Required readings (before class)

To do (before class)

We'll be using OSU's Novus High Performance Cluster: yourONID@novus.dri.oregonstate.edu

You'll need to install two (possibly three) more apps:

  1. CISCO AnyConnect client to connect to OSU's VPN (see download and instructions here);
  2. An sftp GUI (e.g., Cyberduck);
  3. (Windows-users only) An ssh client (see OSU's recommendations, e.g., MobaXterm and PuTTy which also do sftp).

You'll need to login to OSU's VPN before you'll be able to connect to OSU's servers from off-campus. The sftp GUI will make it easier to transfer files (scripts and output) to/from OSU's servers. The ssh client is for interacting with OSU's servers via command-line (for which Mac users can use the built-in Terminal app).

Class structure

  • Interactive computing by command-line
  • Job submission using SLURM

Demonstration files

Serial

Parallel

Additional examples in the examples folder.

Resources

Using the Novus HPC cluster

SLURM explanations

Unix
Additional OSU HPCs (soon to be partly merged)