The ‘Jupyter and Spark’ option provides you with a popularly used default setting for conducting your AI experiments. It combines distributed data processing by nodes (Spark), with a notebook that will allow you to generate code, annotations, visualizations and so on in a web environment (Jupyter).
The ‘Baremetal’ option allows you to have on-demand resources (CPUs, RAM and storage) to carry out your experiments. When filling out your request, please bear in mind that the infrastructure has a limited capacity for both computing and disk space, so please, try to adjust the request as much as possible to your needs.
