Biowulf Computational Cluster
The NIH Biowulf Cluster provides researchers with a world-class system to assist in solving complex biomedical problems as diverse as gene variation in worldwide human populations and PET brain scan processing to identify plaques associated with Alzheimer's disease. The Biowulf cluster requires a Helix Systems account and a justification describing the computational requirements of the proposed research project.
The Biowulf cluster comprises over 18,000 processors operating cooperatively with each other, communicating over dedicated high-speed networks and utilizing shared file servers. Biowulf cluster users have access to a full suite of scientific applications and databases, including tools for sequence analysis, linkage analysis, phylogenetic analysis, microarray analysis, gene regulation and pathways, computational chemistry, molecular modeling, structural biology, mathematical analysis, proteomics, image analysis, molecular graphics, program development, and scientific databases.
Licensed applications include Bioconductor, Gaussian, Mathematica, and MATLAB (including Simulink, Bioinformatics, Curve Fitting, Image Processing, Neural Network, Optimization, Partial Differential Equation, Signal Processing, SimBiology, Spline, Statistics, Symbolic Math, System Identification, and Wavelet toolboxes).
This service is available to all NIH customers.
Frequently Asked Questions
Q: How do I obtain a Biowulf account?
Q: How much does a Biowulf account cost?
A: Biowulf is not paid for directly by users, because Biowulf is supported by the NIH Management Fund. The cost is spread across all of intramural NIH.
Q: How many CPUs are available to a single user?
A: It depends on how busy Biowulf is, but in general a single user is allowed between 512 and 1024 cores.
Q: How can I learn how to use Biowulf?
A: The Helix Systems Staff teach classes at least once a year. Registration for these classes is available at http://training.cit.nih.gov/. We also host informal workshops and clinics on a regular basis.
Up to Top