---
tags: research-computing, Azure
---
# Azure Labs for DES courses

**Basic premise:** Azure Labs provides 'Windows Desktops in the Cloud' with a large number of pre-installed tools and applications suitable for machine learning and deep learning courses.

There is a cost associated with this, so it is important that academics are guided to make the right decisions.

There are three sizes of student machine (Virtual machine or VM):

* Small (2 cores and 4GB memory) <= TRIAL
* Medium (4 cores and 8GB memory) <= GENERAL MACHINE LEARNING
* GPU (Small GPU, 7 cores and 16GB memory) <= DEEP LEARNING

## What information is needed for Azure Labs setup?

*Realistically need this 3 wks before start of module*

Submit all the information as a ticket: https://bit.ly/arc-help

MS DSVM docs: https://docs.microsoft.com/en-us/azure/machine-learning/data-science-virtual-machine/overview

1. Name of the module
2. Module identifier
3. Responsible academic (to sign off) <= 2 weeks before module start - "I have logged in and I agree this is what I need" (if not, talk to RC)
4. VM size
5. Number of hours to be allocated to each student (notional)
6. Is any shared storage required for (eg) large, common shared datasets (cannot be done retrospectively)? If so, how much capacity?
7. Module start date
8. Module end date (although this also needs to include time for resubmissions etc.)

*Around 1 week before module start*

1. Confirmation of sign-off from module lead
2. List of students (this can be quite late- not too late)

## Communicating connection details to students

*Friday before module starts on a Monday*

1. Must use their UoL identifier
2. Connection and sign-up link in Minerva
3. Basic instructions PDF also in Minerva

## At the end of a module

1. Submit another ticket to IT requesting the deletion of resources

**Undeleted resources continue to be billed to DES**