The Neuroscience Institute (NI) and the Center for the Neural Basis of Cognition (CNBC) at Carnegie Mellon University maintain a dedicated computing infrastructure to support the research needs of faculty, staff, and students. These facilities provide secure access to advanced computing resources, high-speed networking, and a comprehensive suite of software tools essential for modern neuroscience and data-intensive research.
Our server infrastructure is housed in a climate-controlled machine room in the Mellon Institute, with systems protected by redundant power and uninterruptible power supplies (UPS) to ensure high availability. Central to our storage capabilities is a Dell PowerEdge R740XD server running Ubuntu Linux, offering 160 TB of enterprise-grade disk space configured with RAID 6 and the ZFS file system. This system provides automated snapshotting and offsite replication to safeguard research data. An additional 100 TB ZFS-based storage system supports the IARPA-funded MICrONS project, with equivalent data protection and reliability measures.
For high-performance computing (HPC), the NI leverages a secure, two-rack GPU/CPU cluster—known as the MIND Cluster—located in the School of Computer Science (SCS) data center. This professionally managed facility is monitored 24/7 by SCS system administrators in collaboration with NI computing staff. The MIND Cluster provides a powerful and flexible infrastructure designed for large-scale neuroscience computation, artificial intelligence, and data analysis.
As of late 2025, the MIND Cluster consists of 16 compute nodes encompassing 32 processors, 512 physical cores, and 1,024 logical cores, with a combined total of 6.8 TB of RAM. Thirteen of the cluster’s nodes are GPU-enabled, collectively hosting 70 NVIDIA GPUs, including 18 L40S, 4 RTX A5000, 2 RTX 3090, 10 TITAN RTX, 16 GeForce RTX 2080 Ti, and 20 TITAN X (Pascal) accelerators. These GPUs support a wide range of computational workloads, from deep learning and large-scale simulations to high-throughput imaging and data analytics.
Major infrastructure upgrades completed in 2025 expanded the cluster’s capacity and performance. New high-memory compute nodes were added, providing over 3 TB of aggregate RAM and 144 additional physical cores. GPU capacity was enhanced with the installation of nodes featuring up to 1 TB of memory and high-speed NVMe SSDs for local scratch storage. The InfiniBand network fabric was upgraded to 100 Gbps full bidirectional EDR to enable faster data transfers, and several GPU nodes received new 8 TB SSD scratch drives to accelerate I/O-heavy tasks. The cluster’s storage backbone was modernized with 12 Gb/s SAS enclosures and additional drive bays for future scalability. A new head node and dedicated login node were brought online, and the operating system across all nodes was upgraded to Springdale 8 (based on CentOS 8) for enhanced compatibility and long-term support.
The cluster is managed using SLURM (for job scheduling), Ansible (for system configuration), and Hydra (for user environment setup). Data partitions employ ZFS with RAID 6 for high reliability, with incremental backups and replication to secure offsite systems managed by SCS. Researchers access the cluster via SSH and have access to a comprehensive software suite for neuroscience, statistical analysis, and machine learning, including AFNI, Freesurfer, MNE, SPM, Matlab, R, DSI Studio, TensorFlow, and others. Singularity and Anaconda environments are supported for containerized and reproducible workflows.
The Computing Center operates as a 100% recharge facility, meaning user fees cover operational and staff costs, while capital improvements and hardware purchases are typically funded through faculty grants or startup resources. Continued investment from researchers ensures that the NI computing environment remains current and capable of supporting cutting-edge research.
The computing infrastructure is overseen by David Pane, NI’s Manager of Computing, who maintains documentation, provides user support, and manages all aspects of system operation. He assists users with software installation, technical troubleshooting, and guidance related to the computing environment within the Neuroscience Institute.