High Performance Computing Core | Research

The High-Performance Computing Core (HPCC) manages AWS and GCP cloud resources and oversees a 4,000 square-foot, state-of-the-art data center located in the WCM Belfer Research Building (BRB). It also collaborates with WCM-Qatar in Doha to provide high-performance computing services. In addition, the HPCC maintains multiple Globus endpoints for research data transfer and ensures high-bandwidth connectivity to Internet2 and other commercial networks, facilitating collaborations with academic, government, and industry partners. The HPCC also offers mid-tier storage services, allowing researchers to archive unused files efficiently.

Our Services

What's New!

Expansion of High-Performance Computing

AI Cluster

In partnership with Dr. Mert Sabuncu from Radiology, ITS has constructed a cutting-edge high-performance computing (HPC) cluster specifically designed for AI-driven image training. The cluster features high-memory nodes, A100 class GPU servers, a 200GB/s InfiniBand cluster network, and storage tailored to support AI workloads. It is optimized to facilitate large language models (LLMs) across various disciplines.

Cayuga Cluster

In collaboration with Cornell's Center for Advance Computing. This cluster offers A100 and A40 GPUs, high memory nodes and 200GB/s HPC cluster network. The system is coupled with a hybrid (flash/HDD) DDN Lustre storage, scalable to 10PB.

Google Cloud Platform

Google Cloud Platform (GCP) offer robust solutions for a variety of storage and compute needs. These environments are designed to handle large-scale data processing and complex computational tasks, making them ideal for research projects. Sensitive data analysis, including protected health information (PHI), is authorized in AWS and Google Cloud Platform (GCP) environments managed by ITS.

On-premises HPC (High Performance Computing)

All clusters are connected to on-premises storage via Infiniband interconnect to allow fast and reliable access to your data. We can assist with designing custom private clusters for your team.

Computational Service Details

From hot on-premises Lustre store to archival cloud storage

SCU BRB Cluster uses Lustre filesystem on-premises data storage system with over 7 PB usable capacity. Lustre is connected to clients via a 100Gbps Infiniband fabric.

Lustre operates three filesystems and fast local NVMe: /home, /athena, /midtier, and /scratch.

We can also assist with moving data to cloud-based storage providers, offering inexpensive long-term archival storage.

Storage Service Details

Development consulting and resources for software developers

Our team of software engineers consults and assists with your software development needs. We provide on-demand training and consulting, as well as custom software development services.

We also provide resources to assist your own development efforts. We offer source control and continuous integration via our private GitLab server. We also publish best practice guides on various development frameworks.

Development Service Details

Visit our knowledge base for further information on the SCU resources

HPCC Wiki

Get in Touch

We are here to help solve your problem. Please select from these options to find the best way to contact us:

What is the 'SCU Account'?

I'd like to request a new SCU Account.

I need to reset my password.

I have other questions.