Department of Systems Biology Computing
The HPC cluster at the Department of Systems Biology Information Technology (DSBIT) is a world-class academic computing environment that provides powerful and secure compute capabilities and personalized support.
Available Hardware
Housed in two data centers totaling more than 3,000 sq. ft. of floor space, HPC at DSBIT is Linux-based (CentOS 7) and contains 6,384 total compute cores, 23TB of total RAM (random access memory) and 148 NVIDIA GPU providing an additional 75,776 total CUDA cores .
● 528 HP blade systems
○ Each node has 12 cores (2 hex-core processors)
○ 480 nodes with 32GB memory
○ 50 nodes with 96GB memory
● 2 large (1TB) memory servers
● 2 head nodes and a pool of virtual login nodes
● 10 Gbps Ethernet connection to the cluster network
● 32 nodes linked with 40 Gbps QDR InfiniBand
DSBIT follows strict best-practices data-storage protocols, including:
● Compliance with Columbia University Information Technology (CUIT) standards
● HIPAA-compliant security measures
● Regular data snapshots, replication, and offsite backup
Available Software
The cluster supports a wide range of software, including:
● Latest GNU and Intel compilers for C, C++, and Fortran
● Perl interpreters
● Java SDKs
● Matlab
● R
● Python
● BLAST
● CUDA
● Qiime
● Ruby
Support
Current users can contact the HPC liaison directly with questions or support needs. Please include a full description of your issue along with the submitted script and any error messages you’ve received.
If you have any questions or need further information, contact Rebecca Yohannes, HPC liaison, at ry111@cumc.columbia.edu.