Lunch bytes #3: Accelerated Machine Learning

9 December 2020 - 12:00-13:00
Google meet
Phil Tooley (NAG), Filippo Spiga (Nvidia), NAG and Nvidia

LunchBytes is a new monthly series of short talks for those in the research community at TUOS who work with/write code, use/manage research data and use/manage research infrastructure. We hope through these talks we will come together as a community to discuss best practices and useful methods/tools.

At this LunchBytes event we’ll be exploring the process of accelerating machine learning on the cloud and using machine learning to accelerate physics simulations.

Scalable ML in the Cloud - First Steps on MS Azure

Phil Tooley, HPC Analyst, NAG

For new users the cloud can be a complex and overwhelming environment, even for experienced users of traditional HPC and on-premises computing. In this talk I will look at some of the common issues experienced when deploying machine learning training in the cloud, and how to get going quickly and efficiently. I will share my own experiences of performing distributed ML training on Microsoft Azure, including what worked, what didn’t, and what I wish someone had told me before I started!


Filippo Spiga, EMEA HPC Developer Relations, Nvidia

Accelerating physics and engineering using Physics Informed Neural Network (PINN)

Q & A Jamboard

Google JamBoard for Q&A: note down questions/comments (using JamBoard Sticky Notes) before/during the event.

Joining instructions

This lunchBytes session will be run at the following Google meet ( which is open to members of the University of Sheffield. If you would like to attend and do not have a Sheffield account, please contact the organisers by email at

Contact Us

For queries relating to collaborating with the RSE team on projects:

To contact the RSE team about seminars, training or JADE:

Join our mailing list so as to be notified when we advertise talks and workshops by subscribing to this Google Group.

Queries regarding free research computing support/guidance should be raised via our Code clinic or directed to University central IT support.