TACC Projects
TACC staff are involved in a wide range of research & development projects aimed at improving the performance of our systems and software, from libraries to pool management to next-generation visualization tools. A selection of these projects are listed below. To learn more about each project, please click on the project title.
The list of software that is available for use on TACC's systems is available on the Software Page.
Computational Biology
The Computational Biology Program at TACC is committed to providing the life science research community with computational tools and expertise needed to address modern biological research questions. The TACC team is focused on ensuring that TACC maintains the hardware, software and domain expertise to support a wide variety of bioscience research.
Data Management & Collections
The Data Management & Collections (DMC) group works to meet the needs of faculty and researchers for data collection services, and to contribute to the potential of data-driven research to make discoveries. The group builds and maintains large data-management and storage resources and consults with collections' creators in all aspects of the data lifecycle, from creation to long-term preservation and access. The DMC group actively seeks out research and grant proposal collaborations with researchers and institutions with collections of interest.
Digital Humanities
The proliferation and accessibility of massive online databases of textual, visual and aural resources have brought new complexity to the study of history, languages, civilizations, and human activities. TACC is working with humanities researchers at The University of Texas at Austin to discover new ways that advanced computing can be applied to emerging computational disciplines.
DisplayCluster
DisplayCluster is a software environment for interactively driving large-scale tiled displays. The software allows users to interactively view media such as high-resolution imagery and video, as well as stream content from remote sources such as laptops / desktops or high-performance remote visualization machines.
EnVision/Longhorn Portal
The TACC team created the EnVision guided visualization system, which is currently deployed on Longhorn and available to users through the Longhorn Visualization Portal. EnVision is intended for people with little–to-no visualization experience, providing users with a full set of visualization methods through a simple wizard-based interface.
Flame Iterative
Derivation of Krylov space methods (Conjugate Gradients, GMRES, BiCGstab, et cetera) has long been a dark art. The road from a simple idea to a practical code is long and arduous, and uniqueness and optimality of the endpoint is not guaranteed. We try to remedy this situation by offering a formal treatment that requires a researcher to specify a very high-level predicate describing the operation. After this, the actual derivation will be mechanical and derived in such a way that the result is proved correct.
FutureGrid
FutureGrid is a project to develop a high-performance grid test bed that will allow scientists to collaboratively develop and test innovative approaches to parallel, grid, and cloud computing. The test bed will be composed of a high-speed network connected to distributed clusters of high-performance computers, including a new system at TACC.
GDBase: An Engine for Scalable offline Debugging
Debugging at large scale is a constant challenge in the operation of leadership class systems. GDBase incorporates a lightweight, offline architecture for data collection inspired by performance analysis tools, a database repository for vast quantities of debugging information and cross-run analysis, and an API to create new agents for providing a variety of interface and analysis tools. GDBase lifts productivity across all domains that develop applications for large-scale cyberinfrastructure.
GotoBLAS2
GotoBLAS2 was released by the Texas Advanced Computing Center as open source software under the BSD license. This product is no longer under active development by TACC, but it is available to the community to use, study, and extend. GotoBLAS2 uses new algorithms and memory techniques for optimal performance of the BLAS routines. The changes in this version target new architecture features in microprocessors and interprocessor communication techniques. In addition, NUMA controls enhance multi-threaded execution of BLAS routines on node.
GPGPU Computing
GPGPU (general purpose computing on graphics processing units) is a methodology for high-performance computing that uses graphics processing units to crunch data. In 2010, TACC installed Longhorn, a 256- node Dell visualization and data analysis cluster with two powerful NVIDIA GPUs in each node. In addition to its use as a production visualization supercomputer, Longhorn also serves as a research and development platform for investigating how cluster and GPGPU parallel technologies can be effectively hybridized.
Green Revolution Cooling
Computer processors generate massive amounts of heat. The cost and practicalities of cooling HPC centers necessitate new, efficient and sustainable technologies. TACC has partnered with Green Revolution Cooling to explore an innovative new method of server cooling: total submersion in dielectric fluid coolant, also known as mineral oil.
Hadoop Cluster
TACC will augment the existing Longhorn cluster with distributed storage (local hard drives) and open-source "Hadoop" software, which will together enable cloud computing. Faculty, students, and staff will work collaboratively to develop campus-wide cloud computing expertise at UT via a program of research talks, training workshops, course offerings, and online documentation.
iPlant Collaborative (iPlant)
iPlant was developed to create a new type of organization—a cyberinfrastructure collaborative for the plant sciences—that enables new conceptual advances through integrative, computational thinking. The iPlant Collaborative brings together strengths in plant biology, bioinformatics, computational science and high performance computing, as well as innovative approaches to education, outreach, and the study of social networks.
Karnak
Developed by TACC researchers, the Karnak service provides predictions of when the batch schedulers that manage TeraGrid resources will start jobs. Predictions are provided for hypothetical jobs as well as for jobs that have already been submitted.
The National Archives and Records Administration (NARA)
NARA stores and maintains the official records and digital collections of the U.S. government, which reached 12 petabytes in 2010. In response to this data deluge, NARA turned to TACC to develop a strategy for computationally-assisted archiving. By working with TACC, NARA is drawing on the expertise of digital archivists and data experts to come up with new solutions for this pressing problem.
The Center for Predictive Engineering and Computational Sciences (PECOS)
PECOS is a new DOE funded Center of Excellence within the Institute for Computational Engineering and Sciences at The University of Texas at Austin. PECOS brings together an interdisciplinary, multi-university team to develop the next generation of advanced computational methods for predictive multiscale, multiphysics simulations.
PerfExpert
HPC systems are notorious for operating at a small fraction of their peak performance, and the ongoing migration to multi-core and multi-socket compute nodes further complicates performance optimization. The readily available performance evaluation tools require considerable effort to learn and use. As a remedy, TACC has helped to develop PerfExpert, a tool that combines a simple user interface with a sophisticated analysis engine to detect core, socket, and node-level performance bottlenecks in each procedure and loop of an application.
Portal/Application Development
Providing researchers with powerful and intuitive interfaces allows them to focus on scientific exploration rather than technological hurdles. TACC's Advanced Computing Interfaces group evaluates, develops, and supports software-based interfaces and services to simplify access to HPC resources and increase the effectiveness of computational research.
Salsa Adaptive Large-scale Solver Architecture (SALSA)
SALSA is a software project that aims to assist applications in finding suitable linear and nonlinear system solvers based on analysis of the application-generated data. The SALSA system features heuristic decision making based on a database of performance results that tune the heuristics over time.
Scalable Collaborative and Remove Visualization Software (ScoreVis)
This project is a direct response to the need for next-generation visualization tools for terascale/petascale science that run on large-scale HPC platforms and on dedicated graphics clusters. ScoreVis is targeted at large-scale clusters—with or without graphics processing units (GPUs)—and seamlessly scales across platforms.
Sparse Direct Factorizations through Unassembled Hyper-Matrices
TACC advocates a novel strategy for sparse direct factorizations that is geared towards the matrices that arise from hp adaptive Finite Element Methods. In that context, a sequence of linear systems derived by successive local refinement of the problem domain needs to be solved. Thus, there is an opportunity for a factorization strategy that proceeds by updating (and possibly downdating) the factorization.
Texas Test Problem Server
The Texas Test Problem Server is your source for complicated linear systems to be used in numerical linear algebra, graph theory, and/or performance evaluation/optimization research. The server hosts a number of applications that can be controlled through a web interface to deliver problems as large, complicated, and varied as desired.
Topology-Aware MPI and Job Scheduling
HPC scheduling systems are generally not adjusted for the underlying topology of the computers on which they run. As a result, these systems compute less efficiently than an intelligent decision-making system would. Supported by an STCI grant from the NSF, researchers at TACC are trying to endow the Oracle Grid Engine (OGE), which runs on Ranger, with dynamic topology information to improve its performance.
eXtreme Digital (XD) Technology Insertion Service (TIS)
XD TIS began in 2010 and was one of the initial XD awards provided by the NSF. XD TIS has systemic tools in place to identify the most promising technologies to meet evolving demands. These measures will allow TACC and the other partner centers to be better stewards of XD resources and optimize their use by an increasing number of new users in diverse fields of research.