Research INKlings October 2007 Announcement.
The Biomolecular Computing Resource (BCR) is being transformed into a new URRF-supported entity to be called the Computational Biology Resource Center. A major goal of the CBRC will be establishment of a state-of-the-art computational infrastructure for scientists to apply advanced computer algorithms to biological problems.
Toward reaching this goal, the CBRC will purchase and maintain a 16 node 132 CPU computing cluster combined with multi terabit storage capacity. The cluster, currently in the purchasing cycle, will be a LINUX-based system aimed at supporting a host of biodatabases as well as applications in drug discovery, NMR, x-ray crystallography, DNA microarray analysis, bioinformatics, image analysis and molecular modeling.
The cluster hardware will be a vertical stack of 16 compute node Dell model PE1950 computers each with Xeon Dual Quad-Core (a.k.a. Clovertown architecture) 2.33GHz CPUs plus one master node. The master node, a Dell PE2950, will also use Dual Quad-Core Xeon CPUs and will host the job scheduler software that will task the compute nodes.
The CBRC will encourage broad usage of the cluster. The CBRC oversight committee will establish usage policies and will review and approve requests for installation of specific applications.
The CBRC cluster will run open source and commercial software. Commercial software licenses tend to scale with the number of CPUs and are therefore expensive. CBRC will support some commercial software applications approved by the oversight committee.
Another goal of the CBRC is to support software development. CBRC will provide programming tools in the form of compilers, scripting software, web-based interfaces. In addition CBRC will provide user support in the form of level 2 programmers who will aid users for particular applications and or run software for users.
Advanced users expecting to consume large amounts of time will be expected to present proposals for usage to the oversight committee. CBRC will attempt to support all usage up to the point of physical saturation.
OCIO-IS will house and manage the hardware of the system. Software installation will be the level one programmer's responsibility but decisions about which software to install will come through the oversight committee.
The primary strategic aim of the CBRC is to remove computational barriers to research progress. At the same time the CBRC is a first time venture in managing a campus-wide computing resource run by and for researchers. The hardware, software and management will require regular input from the users.
CBRC time line is approximately as follows:
July 2007 Established a working group to make recommendations to the Associate Provost for Research.
August/September 2007 Hardware specifications approved. Funds allocated. Purchasing process begun. CBRC web site design begun.
October 2007
Hardware arrived by October 6. Hardware to be installed and tested. Initial software and databases are to be installed and testing to begin. CBRC Web Site is to be housed on the new hardware.
November 2007 Open for first MUSC users.
Members of the CBRC oversight committee include:
Scott Argraves- Cell Biology & Anatomy
Craig Beeson- Pharmacy / Pharmaceutical Sciences
Mark C. Bulson- Biostatistics, Bioinformatics & Epidemiology
Christopher Davies- Biochemistry & Molecular Biology
Starr Hazard- BCR Library
Josh Spruil- Cell Biology & Anatomy
Xinghua Lu- Biostatistics, Bioinformatics & Epidemiology
