About Research Computing

What is Research Computing?

Briefly

Research Computing refers to the specialized infrastructure, software needs, and personnel that enable advanced computational and data driven research across disciplines.

In Depth

Research Computing is the term used for the people, infrastructure, services, and expertise that enable computation and data intensive endeavors across disciplines. This includes, but goes well beyond, “running jobs on a cluster.”

Listed below are the five major components making up Research Computing at the University of Memphis

Advanced Computational Infrastructure
Research Computing supports an array of resources that enable computational research.

The main resource is our High-Performance Computing (HPC) cluster used for parallel and large-scale simulations and high throughput computing (HTC) for many independent jobs

Data intensive platforms for large scientific datasets

Accelerators (GPUs, specialized hardware)

Research storage and high-speed research networking

Support for several research-funded advanced computing resources in the form of specification, configuration, location (power and cooling) and security/access.

This infrastructure supports numeric intensive and large-scale simulations that cannot be reasonably done on desktops or general IT systems.

Research Specific Software
Researchers require support and maintenance for many different software needs on the cluster, including:

Compilers, math libraries, MPI, CUDA

Domain specific scientific codes

Data formats (e.g., HDF5)

Workflow tools and schedulers

User-created and open-source research software

Researcher-Facing Services and Expertise
Research Computing relies on team members with experience in HPC usage as well as domain specific knowledge to help guide researchers within their own fields. This expertise contributes to:

Consulting on code performance, scaling, and parallelization

Helping researchers choose appropriate computational methods

Supporting reproducibility and research workflows

Advising on data management and compliance

Training faculty, postdocs, and students in computational methods

Research Computing functions as a service organization in support of the research mission. As research grows, discipline-specific expertise in the sciences, engineering, finance and more is needed in support of that growth.

Enabling Research Across Disciplines
Research Computing intentionally seeks to supply resources across different research domains:

STEM fields (physics, chemistry, engineering, earth science)

Life sciences (biology, genomics, neuroscience)

Financial sciences (market analysis)

Social sciences (large scale simulations, network models)

Digital humanities (text analysis, image processing)

Research Computing emphasizes support for a broad, heterogenous range of scientific research and diverse computational models; it does not focus on a single discipline or workload type.

Governance, Security, and Sustainability
Research Computing also guards their researchers from undesirable research conditions like unnecessary wait times and unsafe global communication by enforcing cluster-wide operation guidelines:

Allocation and fair share policies for resource access

Compliance with security and funding requirements

Long term forecasting and sustainability of research platforms

Secure participation in the broader national and international research community

National HPC standards explicitly recognize that research computing environments have different security, lifecycle, and risk profiles than enterprise IT systems.

Here are a few examples of how Research Computing differs from general IT

General IT	Research Computing
Email, ERP, desktops	HPC, computation and data intensive platforms
Standardized software	Highly customized research software
Predictable workloads	Bursty, experimental workloads
Service stability focus	Research enablement focus

These distinctions are why Research Computing is treated as a separate function from central IT, even when organizationally adjacent.