The Multiscale Computing Laboratory
The Multiscale Computing Laboratory is the area of the Department of Biomedical Informatics focused primarily on software for enabling efficient scientific data sharing on the Grid.
The goals of the multiscale computing laboratory are to develop the middleware technology and techniques to enable management, sharing, and manipulation of data at multiple scales across heterogeneous, dynamic collections of storage and computation systems. Some application areas include:
- large scale collaborative biomedical clinical studies
- analysis of gene expression and functional imaging information
- imaging, analysis and simulation of oil reservoirs and data driven control of oil production
- analysis of satellite data
- analysis of multi resolution, multiple-grid simulation datasets
The lab targets techniques to support optimized distributed data storage, indexing, retrieval and processing of large datasets distributed across many distributed storage systems. Its approach is to develop systems software able to leverage knowledge of descriptive metadata in a way that supports broad ranges of application areas. The lab's efforts aim to support analysis, associative query, and visualization of very large datasets distributed among storage systems located in multiple machines and clusters. It develops methods to optimize the performance of spatio-temporal and relational queries directed against these large grid-based, federated datasets and targets multiple-query optimization techniques, along with Grid-based semantic caching and retrieval.
Overview of Current Computing Resources
The MSC Lab has a number of computer clusters that it uses for research and testing purposes. A NSF research infrastructure grant was recently awarded to BMI, the Ohio Supercomputing Center (OSC), and the OSU Department of Computer Science and Engineering (CSE). The lab will leverage this cutting edge compute, memory, and storage system, in addition to our local clusters. Below are the technical specifications for the clusters run by BMI and the MSC Lab.
- OSUMed Cluster (located at the State of Ohio Computing Center (SOCC))
- 1 login node: Dual Pentium III 930 MHz, 512 MB RAM
- 24 computer nodes: Pentium III 930 MHz, 512 MB RAM per node
- Approximately 8.4 TB hard drive space (aggregate)
- DC (DataCutter) Cluster
- 5 computer nodes: Dual Intel Xeon 2.4 GHz, 2 GB RAM per node
- Approximately 1.5 TB hard drive space (aggregate)
- Mob (Mobius) Cluster
- 8 computer nodes: Dual AMD Opteron 240, 8 GB RAM per node
- Approximately 12 TB hard drive space (aggregate)