Skip Ribbon Commands
Skip to main content
  
  
Computational Constraints: interconnect latency and/or throughput, data access
Parallelism: <256
Computational System Hosting: University central or research IT, RENCI
Data types: files, SQL database, Hadoop
Data types primary: No
Data Hosting: Renci
Data limitations: Volume of data is very large
Capacity limitations: on campus, between regional and national
Software Defined Networking (SDN): Yes
SDN Description:
Regional and National resources:
Connections to other entities:
  
The work is basic research in the area of chemical catalysts. In short,the work seeks to develop novel quantum mechanical (QM) methods ( as parallel applications) and enable high throughput analysis using workflows containing these applications.
Computational Constraints: number of cores, number of ensemble runs that can be completed
Parallelism: >10k
Computational System Hosting: Lab, Department, National
Data types: files, Parallel file system
Data types primary: files
Data Hosting: Department, National
Data limitations: Volume of data is very large, The approach is new and multidisciplinary. Thus no data-model has been defined yet.
Capacity limitations: bandwidth capacity is not a problem
Software Defined Networking (SDN): Do not know
SDN Description:
Regional and National resources: DOE Lab
The workflows could leverage OSG and XSEDE for some aspects of the work
Connections to other entities:
  
Collaboration environments spanning CASC member campuses, and support for student digital libraries. A data grid can be installed that promotes collaboration, implements NSF style data management plans, preserves reference collections, and enables students to implement their own digital libraries.
Computational Constraints: data access
Parallelism: N/A
Computational System Hosting: University central or research IT
Data types: files, SQL database
Data types primary: No
Data Hosting: University central or research IT
Data limitations: collaborative research on shared data
Capacity limitations: on campus
Software Defined Networking (SDN):
SDN Description: Policy based systems can be implemented that minimize impact on a network, through migration of analysis to storage systems.
Regional and National resources: XSEDE, DataNet project
Interoperability with existing infrastructure. Data grids manage the multiple name spaces needed within a collaboration environment (user names, file names, workflow sharing, metadata, policies, procedures, and storage repository names). For each name space, data grids define the operations that are supported. Virtualization mechanisms are then implemented to ensure the operations can be applied across existing infrastructure.
Connections to other entities: >50
  
At RENCI, we are working with the UNC Medical School to develop the informatics infrastructure that can present clinicians with the clinically relevant mutations identified in a patients genome. The system include HPC workflows that are used to align genomic reads and identify genetic variations, database and Hadoop based storage systems that provide data management and annotation of variations based on clinical relevancy, and workflow systems that orchestrate the processing of the entire process.
Computational Constraints: number of cores, systems for population level, data intensive computations
Parallelism: 512-1024
Computational System Hosting: University central or research IT, National, RENCI
Data types: files, SQL database, Hadoop, Parallel file system
Data types primary:
Data Hosting: University central or research IT, RENCI
Data limitations: Volume of data is very large, Inherent complexity of the data
Capacity limitations: between campus and regional
Software Defined Networking (SDN): Do not know
SDN Description:
Regional and National resources: OSG
Scaling of data, both in terms of the number of terabytes of data managed and in terms of performing data-intensive computations on the data. We lack strong approaches for switching from batch-oriented computations across data sets to continuous computing as new data arrives.
Connections to other entities:
  
DYNES @RENCI
Computational Constraints: data access
Parallelism: N/A
Computational System Hosting: Lab, Department, University central or research IT
Data types: files, SQL database
Data types primary: files
Data Hosting: Lab, Department, University central or research IT
Data limitations: Volume of data is very large
Capacity limitations: on campus, between campus and regional, between regional and national
Software Defined Networking (SDN): Yes
SDN Description: SDN provides the potential to control/influence network resources and how they're utilized. At an application level, SDN provides an interface to optimize how applications use the network, computation, and storage, which ultimately enables tighter integra
Regional and National resources: DOE Lab
SDN provides the potential to control/influence network resources and how they're utilized. At an application level, SDN provides an interface to optimize how applications use the network, computation, and storage, which ultimately enables tighter integration between these resources and better application performance.
Connections to other entities: 20-50
  
The fusion process in which two bilayers merge to form a continuous structure, is influenced by many factors including composition, hydration, electrostatics, and environmental conditions. Fusion of lipid vesicles is essential for the life cycle of all living organisms. In industrial processes, however, vesicle fusion is an undesirable event leading to product instabilities. A number of formulated consumer products are made of dispersed vesicles which contain active components. The purpose of this project is to elucidate the molecular picture of these processes and how they can be controlled through changes of the environment or composition.
Computational Constraints: number of cores, interconnect latency and/or throughput
Parallelism: >10k
Computational System Hosting: Lab, University central or research IT, National
Data types: files, Parallel file system
Data types primary:
Data Hosting: Lab, University central or research IT, National
Data limitations: Data analysis tools not designed for size and nature of problem
Capacity limitations: between campus and regional, between regional and national
Software Defined Networking (SDN): Do not know
SDN Description:
Regional and National resources: XSEDE, DOE Lab
There is not enough compute capacity available on medium to high capability resources. Sufficient capacity is already currently only available by using high end capability resources at the as high capacity level and as low capability level as tolerated by the operating provider and necessary for sufficient project throughput. Competition for these resources is strong and expected to increase.
Connections to other entities: