Skip to content

Jupyter At FCC AFs

One of the services available at the FCC analysis facilities is the Jupyter Lab environment.

Table of Contents

Accessing the JupyterHub

The BNL JupyterHub is

https://jupyter-fedid.sdcc.bnl.gov/

Before accessing it, please apply for a BNL computing account. Link to the instructions.. BNL's Scientific Computing and Data Facility (SCDF) provides JupyterHub environment on their HTC (high throughput computing) cluster and HPC (high performance computing) clusters. Choose one of them to login.

The JupyterHub resources on HTC cluster and HPC clusters

Jupyter instances running under the HTC JupyterHub are actually Condor batch jobs. Using HTC JupyterHub requires affiliation to an experiment, such as FCC.

Jupyter instances running under the HPC JupyterHub are run as SLURM batch jobs on the BNL Institutional Cluster (IC). Using HPC JupyterHub requires a valid time allocation on those clusters. Nvidia GPUs are available on the IC cluster.

Kernels and extensions in the FCC Jupyter environment

The Jupyter environment provides several kernels and extensions. This includes:

  1. python3 kernel with the Key4hep software stack, which includes the FCC software stack.
  2. python3 with pyroot and uproot.
  3. python3 with ML packages such as Tensorflow and Keras, with support of Nvidia GPUs.
  4. Terminal console for simple interactive use, e.g. file managements.
  5. Terminal console for simple interactive use
  6. Markdown document editor and previewer. You can edit and preview in two tabs simultaneously.

Getting help

Need help? Have questions or comments?, Visit our Discourse Logo AF Discourse Forum (do not confuse with Discord 👾) for user support, contact, friendly discussion, newsletter and more! We'd love to help you have a smooth experience while working at our analysis facilities! Note that there is a specific "BNL FCC analysis facility" section for questions, but the discourse forum is shared as the FCC AF shares many technical setups with the ATLAS AF.