From UABgrid Documentation
HPC Web Portal now in Beta
The new HPC web portal is now available. We encourage you to try it out as an alternative to traditional clients. It provides file, shell, and desktop access to the cluster within your web browser.
Welcome to the Research Computing System
The Research Computing System (RCS) provides a framework for sharing data, accessing compute power, and collaborating with peers on campus and around the globe. Our goal is to construct a dynamic "network of services" that you can use to organize your data, study it, and share outcomes.
'docs' (the service you are looking at while reading this text) is one of a set of core services, or libraries, available for you to organize information you gather. Docs is a wiki, an online editor to collaboratively write and share documentation. (Wiki is a Hawaiian term meaning fast.) You can learn more about docs on the page UnderstandingDocs. The docs wiki is filled with pages that document the many different services and applications available on the Research Computing System. If you see information that looks out of date please don't hesitate to ask about it or fix it.
The Research Computing System is designed to provide services to researchers in three core areas:
- Data Analysis - using the High Performance Computing (HPC) fabric we call Cheaha for analyzing data and running simulations. Many applications are already available or you can install your own
- Data Sharing - supporting the trusted exchange of information using virtual data containers to spark new ideas
- Application Development - providing virtual machines and web-hosted development tools empowering you to serve others with your research
Support and Development
The Research Computing System is developed and supported by UAB IT's Research Computing Group. We are also developing a core set of applications to help you to easily incorporate our services into your research processes and this documentation collection to help you leverage the resources already available. We follow the best practices of the Open Source community and develop the RCS openly. You can follow our progress via the our development wiki.
The Research Computing System is an out growth of the UABgrid pilot, launched in September 2007 which has focused on demonstrating the utility of unlimited analysis, storage, and application for research. RCS is being built on the same technology foundations used by major cloud vendors and decades of distributed systems computing research, technology that powered the last ten years of large scale systems serving prominent national and international initiatives like the Open Science Grid, XSEDE, TeraGrid, the LHC Computing Grid, and caBIG.
The UAB IT Research Computing Group has collaborated with a number of prominent research projects at UAB to identify use cases and develop the requirements for the RCS. Our collaborators include the Center for Clinical and Translational Science (CCTS), Heflin Genomics Center, the Comprehensive Cancer Center (CCC), the Department of Computer and Information Sciences (CIS), the Department of Mechanical Engineering (ME), Lister Hill Library, the School of Optometry's Center for the Development of Functional Imaging, and Health System Information Services (HSIS).
As part of the process of building this research computing platform, the UAB IT Research Computing Group has hosted an annual campus symposium on research computing and cyber-infrastructure (CI) developments and accomplishments. Starting as CyberInfrastructure (CI) Days in 2007, the name was changed to UAB Research Computing Day in 2011 to reflect the broader mission to support research. IT Research Computing also participates in other campus wide symposiums including UAB Research Core Day.
Featured Research Applications
The Research Computing Group also helps support the campus MATLAB license with self-service installation documentation and supports using MATLAB on the HPC platform, providing a pathway to expand your computational power and freeing your laptop from serving as a compute platform.
The UAB IT Research Computing group, the CCTS BMI, and Heflin Center for Genomic Science have teamed up to help improve genomic research at UAB. Researchers can work with the scientists and research experts to produce a research pipeline from sequencing, to analysis, to publication.
Grant and Publication Resources
The following description may prove useful in summarizing the services available via Cheaha. Any publications that rely on computations performed on Cheaha should include a statement acknowledging the use of UAB Research Computing facilities in your research, see the suggested example below. We also request that you send us a list of publications based on your use of Cheaha resources.
Description of Cheaha for Grants
UAB IT Research Computing maintains high performance compute and storage resources for investigators. The Cheaha compute cluster provides over 2800 conventional INTEL CPU cores and 80 accelerators (including 72 NVIDIA P100 GPUS's) interconnected via an InfiniBand network and provides 468 TFLOP/s of aggregate theoretical peak performance. A high-performance, 6.6PB raw GPFS storage on DDN SFA12KX hardware is also connected to these compute nodes via the Infiniband fabric. An additional 20TB of traditional SAN storage is also available for home directories. This general access compute fabric is available to all UAB investigators.
Acknowledgment in Publications
This work was supported in part by the National Science Foundation under Grants Nos. OAC-1541310, the University of Alabama at Birmingham, and the Alabama Innovation Fund. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the National Science Foundation or the University of Alabama at Birmingham.