Recent advances in neuroimaging have greatly enhanced understanding of the brain and mind. With these developments come continually growing computational demands and challenges. From increases in the sheer volume of data due to improvements in spatial and temporal resolution, to the complexities of integrating neuroimaging data sets derived from multiple imaging modalities with behavioral and genetic information, the computational requirements for teaching, data analysis, storage, visualization, sharing/availability and security demand ever advancing resources and expertise.
The Neuroimaging Compute Facility (NCF) is a central enabling infrastructure for neuroimaging teaching and research whose mission is to provide high performance, high power, robust, reliable and secure computer systems and human expertise to meet the challenges of neuroimaging research and teaching. The NCF is available to all Harvard investigators collecting MRI data in the Neuroimaging Facility and is managed by FAS Research Computing, a team of 15+ full-time technical professionals (http://rc.fas.harvard.edu). Students and researchers interested in using the NCF must sign up for a User Account. Instructions for requesting an account are here.
For answers to frequently asked questions about using the NCF, click here for our FAQ page.
The NCF data processing cluster consists of 12 Dell PowerEdge blades (8 M915 blades with 64 core AMD Opteron™ Processor 6274 with 256 GBs of RAM, 4 C6145 blades with 64 core AMD Opteron™ Processor 6274 with 256 GBs of RAM) and 1 ‘high memory’ node (Dell PowerEdge R930 with 40 cores and Intel® Xeon® CPU E7-8891 v4 with 2.8GHz and 3TB of RAM). This is a total core count of 808 with a combined 6TB of RAM.
Data Storage (CBS Central):
All nodes in the cluster share a pool of network-attached storage that has a current capacity of 600TB and a potential storage growth of 1PB. All systems are tied together via a 4Gbps TCP. In addition, data is continuously backed up, 100 miles away, at the Massachusetts Green High Performance Computing Center in Holyoke Massachusetts.
This IT infrastructure is a separate, secure private RFC1918 compliant network isolated from the rest of the university through layers of access controls consistent with Harvard University security policies. Only vetted users are given accounts. User passwords are always encrypted when traversing the network. The only external entry points to the network are via an SSH gateway and Virtual Private Network (VPN) connection requiring two-factor authentication. Connectivity to the cluster and storage is provided via 10G links on an enterprise class Cisco switch.
Common tools for data analysis are available and updated regularly. Licenses and software packages for data analysis include, MATLAB, FreeSurfer, FSL, AFNI, and SPM. Significant internal software development for image and data analysis as well as informatics (e.g., XNAT) is ongoing by the Harvard Neuroinformatics Research Group (http:\\neuroinformatics.harvard.edu). This shared resource includes common processing scripts and tutorials for neuroimaging data analysis, including merged atlases for analysis of older adults and children, as well as innovative procedures for high resolution data analysis and surface visualization. To access these software tools, register for an NCF computer account.
– The eXtensible Neuroimaging Archive Toolkit (XNAT)
The XNAT database provides infrastructure for neuroimaging data management and data sharing (www.xnat.org).
– Open Access Series of Imaging Studies (OASIS)
Open access data for teaching and research: Freely available data provides opportunities for those learning to explore analysis tools and become familiar with neuroimaging research. Such data are also important to methods development and novel discovery. A continually growing data library is provided for the Harvard community and other interested individuals (www.oasis-brains.org).
For assistance with IT related issues (including registering your computer with the Harvard network, workstations, wireless concerns etc.), or questions about using the NCF, contact: email@example.com
Please provide as much detail as possible so that the RC staff can triage your problem quickly.