NU Research Computing

Practical, reviewed guidance for running research workloads on the Irgetas, Shabyt and Muon clusters.

Start a first job Choose a system

New to HPC? Connect, transfer files and submit a small Slurm job.

Run a workload Use tested serial, OpenMP, MPI, array and GPU patterns.

Find software Work with Lmod, Conda, compilers, CUDA and applications.

Check limits Review partitions, quotas, wall times and data responsibilities.

Configuration changes over time

Treat live Slurm and Lmod output as authoritative. Before a large run, check sinfo, scontrol show partition, module spider, and the policy page.