Skip to content

SCRC vs NYU HPC

This page provides guidance to help you decide whether to use the computing resources provided by the NYU Stern Center for Research Computing (SCRC) or the central NYU High Performance Computing (HPC) facilities.

Overview

Both SCRC and NYU HPC offer powerful computing resources, but they differ in scope, scale, software availability, and support models. Choosing the right resource depends on your specific research needs, computational requirements, and familiarity with HPC environments.

Key Considerations

Here's a breakdown of factors to consider when making your choice:

  • Scope and Scale:

    • SCRC: Provides a moderately sized HPC cluster primarily for the Stern community. It's well-suited for many research tasks within Stern.
    • NYU HPC: Serves the entire university community with a significantly larger and more powerful cluster, recognized as one of the top academic HPC facilities. It offers extensive capabilities for very large-scale computations.

  • Getting Started & Support:

    • SCRC: Recommended as a starting point, especially for those new to HPC. The SCRC team offers personalized assistance to help Stern users get started and troubleshoot issues.
    • NYU HPC: Offers support to the entire NYU community, but the support model might be less personalized compared to the dedicated Stern SCRC team.

  • Computational Requirements:

    • SCRC: Suitable for relatively modest computational requirements.
    • NYU HPC: Better equipped for very large-scale simulations, data analysis, and tasks demanding massive parallelism or specialized hardware.

  • Response Time / Job Turnaround:

    • SCRC: Typically offers a faster response time for job scheduling and execution compared to NYU HPC, although this can vary depending on current system load.
    • NYU HPC: Due to the larger user base and potentially larger jobs, turnaround times might be longer, though this fluctuates.

  • Software Availability:

    • SCRC: Offers a curated selection of software commonly used within the Stern research community.
    • NYU HPC: Provides a wider range of software options catering to diverse disciplines across the university.

  • Specialized Services (e.g., Big Data):

    • SCRC: Does not have dedicated clusters for services like Hadoop.
    • NYU HPC: Recommended if you need specific Big Data services like Hadoop, as they operate dedicated clusters (e.g., Peel cluster) for these purposes.

Recommendations

  • Choose SCRC if:

    • You are new to High-Performance Computing.
    • Your computational requirements are relatively modest.
    • You are part of the Stern community and prefer personalized support.
    • Faster job turnaround time is a primary concern (though this can vary).
    • The software you need is available on the SCRC cluster.

  • Choose NYU HPC if:

    • You require very large-scale computational resources.
    • You need access to a wider range of software packages not available at SCRC.
    • Your research involves specific Big Data tools like Hadoop that utilize specialized clusters (e.g., Peel).
    • You are comfortable working within a larger, university-wide HPC environment.

Getting Help

The SCRC staff can help you evaluate your computational needs and guide you in selecting the most suitable facility, whether it's SCRC or NYU HPC. Contact the SCRC team at scrc-list@stern.nyu.edu for consultation.