Reference Guide

38 Dell EMC Ready Solutions for AI Deep Learning with NVIDIA | v1.0
5 The Data Scientist Portal
The data scientist portal was developed by Dell EMC and it simplifies the users allowing model
implementation, training and inference tests to be run in Jupyter Notebook or directly from the Linux terminal.
The Jupyter Notebook allows a user to write explanatory text and intersperse it with raw codes and the tables
and figures that those codes generate. It allows a user to directly execute the codes that are embedded in the
document that is being created or has been created. The portal also allows multiple users to share the same
resources in the cluster. In the current version, the resources allocated by the portal is limited to only one node.
If the user needs to use resources that span across two or more nodes, the Slurm job scheduler as described
in Section 5.3 can be used instead. A deeper dive into the portal and its components is listed in Section 2.7.
5.1 Creating and Running a Notebook
A user can access the portal via a web browser on port 8000 (Figure 25). The user can specify the address
without logging into the cluster. The username and password are the same as those for logging to the cluster.
Figure 25: Portal login screen
After logging in, the page is forwarded to the server page (Figure 26). It lists the available server.
Figure 26: Portal server page