Name: John Noss
Job Title: Research Computing Associate
How long have you worked for RC?
A little over two years, I started at RC in August 2012.
What led you to a career in HPC?
I studied physics as an undergrad, and worked in an experimental physics lab, so I was excited to find something that combined being a systems admin with working with researchers on science – whether that’s an experimental apparatus in the lab that’s writing data to RC storage, or computation on the cluster.
What’s the best part of your job?
Definitely the scale of our operation, and the amazing team that we run it with! The Odyssey cluster is a lot of physical infrastructure, but the fact that we can manage the whole process - from racking the servers to provisioning and training, means that even though we’re a small team, we can get a lot of amazing work done.
What’s the hardest part of your job?
That same large scale at which we do things: especially because our team is so small, we have a lot of ground to cover. There’s such a range of science and research that’s being done on the cluster! There are many different software applications being run on the cluster, and each use case is different.
What’s the biggest misconception about RC or HPC in general?
For HPC, I’d reiterate that it’s not just for the hard sciences! There are a lot of things you can do in any field, from text parsing to image processing and more. For RC, don’t hesitate to ask us any question, either how to fix something or just for pointers to general training. We have office hours every Wednesday afternoon at our offices, and we’re very approachable and happy to help however we can!
Given all the research conducted on RC’s Odyssey cluster, is there any one project that stands out for you?
Having worked in an experimental physics lab, I’m partial to anything that’s hooked to an instrument - whether it’s a cosmic-ray detector in the 38 Oxford Street high bay, or a DNA sequencer in the System Biology core, I’m constantly impressed by the many ways that data gets collected. The great part about RC is that we can support the whole pipeline of data, from the instrument to the cluster.
If you could give RC users one piece of advice what would it be?
Learn to use git, github, or any version control system – it’s not just for code. This helps you (and your collaborators) manage your files, including your code or text-based data, and protect yourself from accidental deletions or inadvertent changes. Don’t be afraid to read the man pages or documentation, or ask us!
Desktop, Laptop or Tablet?
Laptop: you need to be able to take it with you, there’s just no substitute for a keyboard.
Copyright © 2015. All Rights Reserved.
Information about how to reuse or republish this work may be available at Attribution.