Rivanna, launched in Fall 2014, is the Cray CS300, a 4800-core, high-speed interconnect cluster, with 1.4 PBs of storage. It is composed of 240 compute nodes, each with two 10-core processors and FDR Infiniband interconnect along with a parallel filesystem capable of providing about 25Gb/sec bandwidth. The Cray cluster combines large amounts of processing with large amounts of memory to provide a significant new resource for computationally-intensive research at U.Va. In Spring 2015, the HPC computing resources previously available in Fir were consolidated into Rivanna's environment, increasing the Rivanna environment to over 6,000 cores.
To request a startup allocation, complete the application process here. To request larger allocations, researchers should fill out the proposal request forms below:
Read the Rivanna Annual Report, October 2014-September 2015.
UVa's newest cloud infrastructure is a compute environment for highly sensitive, secure data, providing researchers access to greater computational power and memory in a protected environment. Ivy consists of 32 blades, each with 36 cores, along with 2.65 PB of storage in a virtualization environment. Another 30 nodes, each with 20 cores, and another 500 TB of storage will be added in a hadoop environment in a later phase. Ivy will enhance the capacity of U.Va. researchers to access and use growing sources of sensitive data (for example, HIPAA data or data containing personally identifiable information).
Launched in Fall 2015, Ivy is in the early adopter phase. If you're interested in learning more about Ivy, whether this environment is appropriate for your research, contact the Ivy User Support Team at firstname.lastname@example.org.
Training and Support
The Library's Research Data Services offers consultations and coordinates workshops on data analysis and statistics, computational and programming, and GIS methods and software. See the Research Data Services pages for the current offerings.
- Spring 2016: Introductions to R, Python, LaTeX, Linux; Machine Learning, Social Network Analysis, Factor Analysis; Using Python with Web APIs, Git, Computer Networking; and more!
Advanced Research Computing Services (ARCS) offers consultations on computation-intensive projects, provides orientations to Rivanna on Friday mornings, and teaches CS 5014 Computation as a Research Tool. See the ARCS course page for their current schedule.
- Spring 2016: CS 5014: Introduction to Computation as a Research Tool; Introduction to the HPC System (Fridays)
- Spring 2016: Automated Image Analysis with ImageJ and R; Power and Sample Size Analysis; Building 'Shiny' Web Applications with R; and more!
The Data Science Institute, in partnership with the Library's Research Data Services, sponsors 1-credit short courses. The graduate-level short courses meet for 5 weeks and are designed to provide some focused, hands-on, training on different topics and tools. Researchers from any discipline are welcome to participate. DSI is also partnering with the School of Architecture to open the 3-credit Data Visualization course to a broader audience.
- Spring 2016: DS 6559-001 Data Wrangling in R; DS 6559-002 Applied Causal Inference (using R); SARC 5400-001 Data Visualization