Great Lakes HIgh Performance Computing Cluster

1. Get Duo Duo is required for all services at the University of Michigan. Information about Duo can be found here.
Great Lakes is a campus-wide HPC cluster designed to meet the diverse needs of researchers across the university. It supports a variety of applications, including simulation, modeling, machine learning, data science, genomics, and more. The platform offers a well-balanced mix of computing power, I/O performance, storage capability, and GPUs. It utilizes Slurm as its workload manager, enabling users to work interactively or submit batch jobs.
Open OnDemand offers several interactive applications for a user to use: RStudio, Jupyter Lab and Notebook, MATLAB, Spark and the ability to set up various Linux remote desktops on Great Lakes or Lighthouse. NOTE: When your session is ready to launch, you are reserving resources and those resources will be charged against your slurm account. For each of the interactive apps you’ll need to select the following:
JOB OUTPUT Recall the --output and --error batch file options from the primer. These are important to direct where your results go so that you can review them. If you do not, you may not be able to access them to troubleshoot your work.
After choosing Jupyter Lab or Jupyter Notebook from the Interactive Apps menu you’ll need to specify your desired version, account, hours, cores, partition (standard or largemem), and memory (2GB minimum). You’ll also need to specify the Anaconda Python module you want to use:   Upon selecting “Launch”, your job will be queued on one of your nodes and shown on the “My Interactive Sessions” screen. As soon as the job’s status is “Running”, you can click on “Connect to Jupyter”.
Users interact with an HPC cluster through login nodes. When using ssh via the command line to access any of our clusters you will be placed on a login node.
After choosing MATLAB from the Interactive Apps menu you’ll need to specify your desired version, account, hours, cores, partition (standard or largemem), and memory (4GB minimum). You also need to choose the version of MATLAB you wish to use:  
Open OnDemand is a way for users to access to their HPC resources via a web interface for the Great Lakes and Lighthouse clusters. Start computing immediately. A simple interface makes Open OnDemand easy to learn and use. This includes:
Partition Policies Slurm partitions represent collections of nodes for a computational purpose, and are equivalent to Torque queues. For more Great Lakes hardware specifications, see the Configuration page. Partitions:
We offer three types of Linux remote desktops, depending on the type of work you will be doing.