OrangeGrid vs. Zest: Understanding Your Cluster

Syracuse University operates two computing clusters, each optimized for different types of computational work. Your welcome email specifies which cluster you’ve been assigned based on your research needs.

Your Home Directory: By default, you have a 4TB home directory at /home/netid/. This NetApp-based storage is automatically mounted on all compute nodes when your jobs run. Use it for scripts, code, data, conda environments, and results. You can also create scratch directories within your home (like ~/scratch/) for temporary files during computations.


OrangeGrid (HTCondor) - High-Throughput Computing

Scheduler: HTCondor
Login: its-og-loginX.syr.edu

Best for: Many independent jobs, parameter sweeps, batch processing, embarrassingly parallel workloads

Key Commands:

  • Submit: condor_submit job.sub
  • Status: condor_q
  • Cancel: condor_rm jobid

GPUs: A100, L40S, A6000 (via SUrge)

Examples: Processing thousands of images, Monte Carlo simulations, batch data analysis, Ollama inference

💻 OrangeGrid Code Examples


Zest (Slurm) - High-Performance Computing

Scheduler: Slurm
Login: its-zest-loginX.syr.edu

Best for: Parallel jobs, MPI, tightly-coupled computations, long-running workloads

Key Commands:

  • Submit: sbatch script.sh
  • Status: squeue
  • Cancel: scancel jobid

GPUs: Primarily A40s (via SUrge)

Examples: GROMACS simulations, deep learning training, computational fluid dynamics, molecular dynamics

💻 Zest Code Examples


Want more details? See our complete resource overview for information about all computing resources. For technical specifications, see OrangeGrid Specifications and Zest Specifications.


Quick Command Comparison

Task OrangeGrid (HTCondor) Zest (Slurm)
Submit job condor_submit job.sub sbatch script.sh
Check status condor_q netid squeue -u netid
Cancel job condor_rm jobid scancel jobid
View resources condor_status sinfo
Job history condor_history sacct

What If My Needs Change?

Research evolves! If your computational needs change:

  • 📧 Email us at researchcomputing@syr.edu
  • We can discuss moving you to a different cluster if appropriate
  • We can provide access to both clusters if your work spans different types
  • We’re here to ensure you always have the right tools