Computer Animation

1. Principal Investigator: Gary B. Wilcox

2. Associated Investigators: John Leckenby, F. Talenti

3. Technical Support Contact: Charles Soto

4. Project Abstract:

   The Computer Animation project will investigate new applications of computer animation in a multimedia Web environment in terms of both content creation and consumption. The Web provides an ideal environment for the delivery of computer generated animation as part of larger entertainment, advertising or educational content. Delivering animation over the Web requires working with various browser plug-ins to ensure that the graphics start to display as they are still being downloaded. Animation movies are concise, so total download time, even on complex pages, becomes reasonable. Development of multimedia material in this environment is far from mature, however. Developing animation content specifically suited for delivery in this form is not currently extensively done on Intel platforms.

5. Equipment Status: Just completed installation of equipment.

6. Research Progress.

   Equipment has just been installed and progress on the grant will begin with the semester. Several projects are proposed for the spring semester.

7. Related Web site:
   http://interactive.cocomm.utexas.edu/

Back to Top

Digital Video Telecommunications Research and Teaching

1. Principal Investigator: Al Bovik

2. Associated Investigator: Brian Evans

3. Technical Support Contact: Sanghoon Lee, slee@vision.ece.utexas.edu

4. Project Abstract:

   This project is directed towards exploiting multiprocessor Pentium, WindowsNT, and MMX technology to develop standard-compliant Digital Video Telecommunications technology in networked software. We are currently studying scalable PC-to-PC video teleconferencing. Uncompressed video data rates range from 10-170 Mbits/sec (MPEG-compliant video) to 250 Mbits/sec (digital color NTSC video). Significant technical challenges include: developing streamlined software codecs for high-speed video compression, and motion estimation and compensation; data organization/sequencing; and network traffic protocols to optimize video multisession send/receive.

5. Equipment Status:

   All requested equipment for Year 1 has been received.

6. Research Progress:

   We have just received the requested equipment, so specific progress cannot be detailed at this time. However, we have been actively developing the theory of foveated video compression and network protocols for handling nonuniformly sampled video streams. To see an example of foveated video, see the web site: http://pineapple.ece.utexas.edu/class/Video/demo.html(use Netscape). We are currently studying scalable PC-to-PC video teleconferencing. Uncompressed video data rates range from 10-170 Mbits/sec (MPEG-compliant video) to 250 Mbits/sec (digital color NTSC video). Significant technical challenges include: developing streamlined software codecs for high-speed video compression, and motion estimation and compensation; data organization/sequencing; and network traffic protocols to optimize video multisession send/receive.

Back to Top

Multimedia Delivery Systems
April 1998

1. Principal investigator: Harrick M. Vin

2. Project Objective

   The main objective of this project is to initiate long-term research on digital course creation and delivery. This project was motivated by our belief that the effectiveness of delivering education can be enhanced by: (1) creating a repository of digital modules containing video lectures, notes, presentation slides, simulation packages, animation, images, and exercises; and (2) providing efficient methods for organizing and browsing through this information space.

3. Project Status --- April 1998

   During the past year, we have designed and implemented a initial prototype of InfoWeave, a visual environment for the creation and dissemination of digital educational material over the Internet. The design of InfoWeave integrates and builds upon a large body of research on hyper-media authoring systems, distance education systems, digital library projects, speech recognition systems, network protocols and techniques for streaming video, and multimedia file system design techniques.

   InfoWeave utilizes video as the primary medium for delivering educational material, and uses the associated textual material to search for specific video segments. It organizes the repository of educational material in terms of modules, each encapsulating information on a specific topic at a certain level of detail. Each module consists of digitized video lectures, overhead slides, notes of the topic, algorithm animations, exercises, and a log of discussions among students and faculty clarifying the concepts discussed in the module. To assist students to navigate through this repository, InfoWeave utilizes links to specify the relationship between two data items within and across modules. By supporting various types of links, InfoWeave supports both reader-driven and author-driven navigation. Moreover, it allows users to control their own view of the information.

   Our initial prototype implementation consists of an authoring environment, which facilitates creation of the modules, and a presentation system that supports efficient methods for organizing and browsing through the information space.

   We are in the process of enhancing the current prototype of InfoWeave to include tools for collaborative distance learning. These tools will enable students to collaborate with each other during the learning process as well as to provide feedback to the system so as to enhance its search capability. Finally, to enable the distribution of the educational material as part of a non-networked environment (e.g., CDROM, DVD, etc.), InfoWeave will support a mechanism to create independent hyper-media documents from its repository.

4. Equipment Status

   All requested equipment for Year 1 of the project has been received and is in active use.

5. Concluding Remarks

   We believe that InfoWeave will provide a richer learning environment than is found in today's classrooms, on educational television, or on the World Wide Web. It will also provide a realistic testbed for experimenting with various techniques for distance learning and self-paced education. We anticipate that our experiment with applying this technology in the educational domain will trigger the development of similar multimedia training programs in health-care, manufacturing, and industrial training domains.

6. Related Web Sites

   Distributed Multimedia Computing Laboratory URL: http://www.cs.utexas.edu/users/dmcl

Back to Top

Virtual Museum Project
ProgressReport 02/16/98

1. Timothy Rowe, Project Director

2. Project Overview:

   The Virtual Museum Project aims to train faculty and students in advanced instructional technologies, by developing multimedia highlighting UT¹s finest museums, galleries, and special collections. A network of highly skilled technical staff already in place, will train students, staff, and faculty as they assist them in developing state-of-the-art instructional multimedia. Our instructional materials will be distributed via Internet and CD-ROM, and as interactive museum and gallery exhibits. Building on the experience and success of UT¹s Virtual Campus, our Virtual Museums Project will break new ground in multimedia development in a highly visible way, while training strategically positioned groups of faculty and students.

   Our training focus is the development of premiere instructional materials that utilize the vast holdings of the Archer M. Huntington Art Gallery, the Texas Memorial Museum, and the Harry Ransom Humanities Research Center. The grand technical challenge for the Virtual Museum Project is in digitizing in great detail the many delicate, complex 3-D objects from art and natural history collections. We will use everything from simple scanners to high resolution X-ray CT scanning, laser surface scanning, video digitization, and other techniques will be used to build highly detailed models of the gems of UT's special collections.

   Our High Resolution X-ray CT Lab is a unique facility that was funded by the W.M. Keck Foundation, the National Science Foundation, and the University of Texas. This is the only such facility in any academic institution in the world. It uses a specially designed industrial-grade machine to scan entire volumes of dense and minute natural objects, like fossils and meteorites, for internal inspection and external modeling. The 3-D data volumes that this machine generates preserve unprecedented levels of detail, all of which can be visualized interactively with a growing host of off-the-shelf software. An individual 3-D dataset generally comprises a volume of several hundred megabytes, and we can generate several such datasets per week. We have now begun to train a large group of graduate students to process CT data for their own research projects, and for curricular materials to be distributed across the Web.

3. Progress to Date (02/16/98):

   In December, 1997 we received four dual-processor 300Mhz NT workstations, a quad-processor 200 Mhz NT workstation, and in January we received two 300 Mhz NT workstations. All are now in operation.The dual-processor machines were installed in Center for Instructional Technologies to be used for training in 3-D visualization, image processing, and for Web/multimedia authoring. One workstation is installed at Texas Memorial Museum and the quad-processor machine and one 300Mhz workstation are in operation at the UT High Resolution X-ray CT Lab. On all of these machines, we have installed and begun to use a range of software for 3-D visualization and animation, multimedia authoring, image analysis, and image processing. Student training with the bulk of this software began last semester in a mixed graduate/undergraduate course on image processing, with a focus on tomographic data and 3-D visualization.

   A specific point of note is that we have chosen VoxBlast - a very powerful off-the-shelf 3-D voxel rendering program - as our major 3-D rendering tool. VoxBalst is designed to address multiple processors. We installed VoxBlast on the quad-processor 200 Mhz machine, and it has increased rendering speed for our huge huge datasets by a factor of about 10 over our previous capability. The equipment provided by Intel will enable us to expand the throughput of our operation, in terms of new research products, newly trained students at all academic levels, and some exceptional instructional materials.

   We have now begun to digitize some of the most exquisite and important objects in UT many special collections facilities, including sculpture from the Huntington Art Gallery, antiquities from the Harry Ransom Center for Study of the Humanities, and fossils from the Texas Memorial Museum.

4. Preliminary results of our work can be seen at the following URL's:
   • UT X-ray CT Lab:
     http://www.ctlab.geo.utexas.edu/index.html

   • The Digital Alligator Web Site:
     http://www.ctlab.geo.utexas.edu/dmg/projects/alligator/utgator.htm

   • Thrinaxodon Digital Atlas Web Site:
     http://www.ctlab.geo.utexas.edu/dmg/projects/thrinax/index.htm

Back to Top

Multimedia Navigation Systems
Status Report (January, 1998)

1. Principal Investigator: Russell Pinkston

2. Associated Investigators:
   
   

   Leckenby
   Stone
   Rowe
   Vin
   

3. Technical Support Contact: Russell Pinkston

4. Project Abstract:
   

   Most navigation systems in networked multimedia assume a focused goal-driven searching, rather than the browsing or exploring activity more generally associated with entertainment. Yet even the term "surfing the Web" implies activities more casual than purposeful in nature. These browsing encounters, perhaps akin to channel-surfing on television, will be demanded by consumers as they get access to a greater number and variety of multimedia content sources. Two obvious examples are music and images, whether they be still or moving. While some access of sound and images may be purposeful, much may be akin to a scan button on a radio. This project will implement and investigate, from the perspective of actual users in research settings, a variety of navigating environments for digital music and image digital collections.

5. Equipment Status:

   Received:

   5 cpus (3 dual processor 300MHz Pentium IIs, 1 300MHz Pentium II, 1 Server)

   3 21" displays

   4 sets of Altec speakers

   Not Received:

   Windows NT Server software

   2 additional displays

   Two dual processor systems and one single processor system have been installed in the U. T. Electronic Music Studios, configured, and attached to the network. The server and the other dual processor workstation have been taken to the Center for Advanced Studies in the Arts, but will not be installed until we receive Windows NT Server and 2 additional displays.

6. Research Progress.
   

   PI has received a Summer Research Fellowship from the Center for Advanced in the Arts, which will free him from teaching responsibilities and provide funding for a graduate research assistant who will be assigned to this project. PI has also applied for a University Research Internship, which if funded, would enable the recruitment of a graduate engineering student to participate in the project.

Back to Top

Distributed Visualization
Status Report (June, 1998)

1. Principal Investigator: Chandrajit Bajaj

2. Associated Investigators: Don Fussell

3. Technical Support Contact: N/A

4. Project Abstract:

   The VisualEyes project shall create a distributed infrastructure for parallel analysis and collaborative visualization of multivariate datasets defined on a multi-resolution hierarchy of hybrid mesh types (unstructured, block structured). A primary goal of our system is to handle very large data visualization tasks, defined to be those involving input data sets larger than the collective primary RAM of a cluster of networked PC graphics workstations. We shall develop and implement out-of-core algorithms for progressive and adaptive visualization, working additionally from compressed input data streams. The distributed analysis and visualization tasks shall include accelerated isocontouring of scalar fields (for multiple iso-contour queries), and progressive hypervolume rendering.

5. Equipment Status: Delivery Pending

6. Research Progress:

   • C. Bajaj ``Data Visualization Techniques'', John Wiley and Sons, (1998), To appear.
   • C. Bajaj, V. Pascucci, D. Schikore, ``Fast Isocontouring for Improved Interactivity'' Proc.\ of the ACM Siggraph/ IEEE Symposium on Volume Visualization, ACM Press, (1996), San Francisco, CA, 39 -- 46.
   • C. Bajaj, D. Fussell, ``Hierarchical Image Based Volume Rendering'' Manuscript, 1998
   • E. Camahort, A. Lerios, D. Fussell, "Uniformly Sampled Light Fields" to appear in Eurographics Workshop on Rendering, June 29-July 1, 1998, Vienna, Austria.

Back to Top

Internet City
Status Report (March, 1998)

1. Principal Investigator: Dr. David Maidment (Center for Research in Water Resources)
   

2. Associated Investigators: Dr. Ken Foote (The Department of Geography) and Dr. Barbara Parmenter (Community and Regional Planning Programin The School of Architecture)
   

3. Technical Support Contact: Ty Lehman lehman@mail.utexas.edu 512-471-3111
   

4. Project Abstract:
   

   We conceive of a Web site, Internet City, in which we accumulate map data from central Texas and provide an interactive Web resource. This resource could provide information related to land use planning, geology, geography, and environmental data. This project would use ArcView and ArcInfo to create the GIS maps that would be served up on the Web site. Internet City will share with both the public and industry the research that is done in our departments.

5. Equipment Status: RECEIVED? (YES/NO)
   

   YES	Qty. 1	SINGLE 300 MHz PENTIUM(R) II PROCESSOR W/MMX(TM) TECHNOLOGY ñ 256MB/4GB
   YES	Qty.2	QUAD 200MHz PENTIUM(R) PRO PROCESSOR ñ 512MB/9GB (ìSERVERî)
   YES	Qty. 3	ULTRASCAN 20T TRINITRON MONITOR
   NO	Qty. 1	NETAPPS BOX
   NO	Qty. 1	INTEL SWITCHABLE HUB, 810109

   
   

6. Research Progress.
   

   We only recently received the equipment listed above. Therefore we have started preliminary Web design and have been conducting trial runs of the map server software.

   
   

Back to Top

Virtual Laboratories and Examinations

1. Principal Investigator: John Kappelman, Professor, Department of Anthropology, The University of Texas at Austin

2. Associated Investigators: None

3. Technical Support Contact: John Kappelman, Greg Weiner

4. Project Abstract:

   Advances in 2-D and 3-D digitizing technologies, multimedia software, CPU architecture, and networking speed now permit the development and delivery of truly interactive computer-based multimedia laboratories and examinations in the sciences. Virtual laboratory include the full range of 2-D color images, sound, and 3-D animations, all mastered within an interactive environment that increases the content and broadens the reach of each scientific discipline. Virtual examinations draw from the full range of materials presented in lecture, the readings, and laboratories, and are presented to the student in the form of multiple choice, matching, and interactive problem solving questions. The questions are integrated with 2D images, sound, video clips, and 3D animations. Exams can include "linked and looped" as well as "expert-nested" questions that truly evaluate the exact level of each individual student's expertise. Grades will be automatically calculated and reported as the last screen of the exam for immediate feedback to the student.

5. Equipment Status:

   We received seven Intel computers and one computer server. Out of the seven computers, two have experienced catastrophic failures and we have to date been unable to reformat their harddrives. These two computers may need to be returned to the factory if we are unable to succeed in getting them to reformat.

6. Research Progress:

   a) I presented a computer poster abstract at the American Association of Physical Anthropologists on 2 April 1998 in Salt Lake City on our work with Virtual Examinations. Two of the Intel computers were flown to Salt Lake City for the poster demonstration. The citation is given below:

   Kappelman, J., C. Bramblett, A. Gordon, K. McCardel, T. Ryan, R. Scott, E. Seiffert, and G. Weiner. 1998. Virtual Multimedia Examinations: Integrating images, video, and animations into an interactive testing program. American Journal of Physical Anthropology Supp. 26: 132-133. Abstract.

   b) I recently received a grant from the National Science Foundation (Division of Undergraduate Education, Course and Curriculum Development) to support our future work in building Virtual Examination templates for the sciences. This award is in the amount of $150,000, with an additional $75,093 in matching monies from UT's College of Liberal Arts. This grant will begin 1 June 1998 and brings together the following UT faculty:

   Anthropology:	Dr. John Kappelman, Paleoanthropology and Geology
   	Dr. Claud Bramblett, Primate Behavior
   Astronomy:	Dr. Robert Robbins, Astronomy and Physics
   Chemistry:	Dr. Brent Iverson, Organic Chemistry and Biochemistry
   	Dr. John McDevitt, Inorganic Chemistry
   Geology:	Dr. Timothy Rowe, Paleontology
   Zoology:	Dr. David Cannatella, Zoology
   	Dr. Bassett Maguire, Zoology

   c) The Intel computers were used in the final production , mastering, and testing of the CD ROM, "Virtual Laboratories in Physical Anthropology" (J. Kappelman, Editor) that is being published by Wadsworth Publishing Company. This CD was demonstrated by Wadsworth at their book exhibit at the American Association of Physical Anthropologists national conference from 2-4 April 1998 in Salt Lake City and met with enthusiastic approval.

   d) A beta test version of "Virtual Laboratories in Physical Anthropology" was used this past academic year in my course, "Introduction to Physical Anthropology" here at UT, with copies of the CD placed on close reserve at the library and the SMURF computer lab , as well as being made available to the students.

   e) I have been invited to join a session on "Anthropology Courseware: Teaching with Electronic Media" to be presented at the American Anthropological Association's national meeting in Philadelphia, PA in November 1998. This session will receive funding support from the National Science Foundation.

7. Related Web Sites: Virtual laboratories in physical anthropology

Back to Top

ADVANCED MODELING & SIMULATION

---

Computational Fluid Dynamics AND HPC
(or HYDRA: A Prototype PC Cluster System for Flow and Transport Simulation)

1. Principal Investigator: Graham F. Carey

2. Associate Investigators: None

3. Technical Support Contact: Robert McLay

4. Project Abstract:

   This project involves the investigation of tightly coupled systems of Intel Pentium II PC's for distributed parallel computation. Of particular interest are flow and transport simulation on moderate size clusters. Such systems have sometimes been termed 'BEOWULF' systems and are a subject of recent interest because of the promise of good price-performance through the use of commodity-off-the-shelf (COTS) components. The approach has been facilitated by the development of the MPI message passing library standard and recent operating system enhancements.

5. Equipment Status: We have received and installed this year's allocation and the 16 processor cluster is operational.

6. Research Progress:

   We carried out system tests and ran some code 'kernels' on the cluster system in December and early January. Following that we implemented a parallel finite element analysis code for nonlinear potential problems as a first test case. The parallel approach is based on domain decomposition with parallel conjugate gradient solution of the linear systems arising from a successive approximation algorithm. The program has been tested and is providing correct results. Parallel cluster performance studies will be carried out in mid year. As a second more significant application we are investigating coupled solution of viscous flow and heat transfer. A prototype parallel code with MPI for communication is being implemented on the cluster. This problem will be the main focus of both our applications and performance studies.

7. Project website:
   • http://www.cfdlab.ae.utexas.edu/intel_beowulf
     
   • http://www.cfdlab.ae.utexas.edu/

Back to Top

Deep Submicron Process and Device Modeling and Analysis
Status Report (January, 1998)

1. Principal Investigator: Al F. Tasch

2. Associated Investigators: Sanjay Banerjee

3. Technical Support Contact: Siva Mudanai, siva@roz.mer.utexas.edu, 512-471-1685

4. Project Abstract:

   This project involves model and simulator development, and design and manufacturing sensitivity analysis in silicon integrated circuit technology in two areas: ion implant modeling, and deep submicron MOS device modeling and analysis. The device modeling and analysis is currently focused on modeling quantum mechanical effects in both inversion and accumulation layers, and accumulation layer mobility; and sub-100 nm modeling and analysis with heterojunctions included. The ion implant modeling project is focused on the modeling of both impurity profiles and implant-induced damage profiles.

5. Equipment Status: RECEIVED? (YES/NO)

   YES	Qty. 7	SINGLE 266MHz SCSI PENTIUM(R) II PROCESSOR W/ MMX(TM) TECHNOLOGY - 256MB/4GB,
   NO	Qty. 1	DUAL 300MHZ SCSI PENTIUM II PROCESSOR W/ MMX TECHNOLOGY - 256/4GB, 4/8GB, DAT AND TAPE BACKUP DRIVE, Qty 1 340 0044,
   NO	Qty. 1	DIAMOND FIRE GL 4000 PCI 32BIT 3D 15MB, 628-113, 1
   YES	Qty. 3	DUAL 300MHZ SCSI PENTIUM II PROCESSOR W/ MMX TECHNOLOGY - 256/4GB,
   YES	Qty. 3	4/8GB, DAT AND TAPE BACKUP DRIVE, 340-0044,
   YES	Qty. 2	ETHEREXPRESS PRO/100 LAN ADAPTER -5 PAK, 802199
   NO	Qty. 2	INTEL SWITCHABLE HUB, 810109
   NO	Qty. 8	ULTRASCAN 20T TRINITRON, 320-5396

6. Research Progress.

   In the last several months we have developed and implemented in UT-MiniMOS, models and source code for hole hydrodynamic transport, and quantum mechanical effects in both electron and hole MOS inversion and accumulation layers. We have also developed a simulator, UTQUANT 2.0, which self-consistently solves Schrodinger's equation and Poisson's equation for both MOS inversion layers and MOS accumulation layers. We have proactively transferred User's manuals, model information, and source code for these models to at least 17 individuals in Intel at Hillsboro, Oregon and Santa Clara, California. Also, Intel has adopted our ion implant simulator, UT-MARLOWE, as their simulator in their technology development.

Back to Top

Reservoir Simulation

1. Principal Investigator: Dr. Kamy Sepehrnoori

2. Associated Investigators: Dr. A.D. Hill, Dr. Larry Lake, Dr. Gary A. Pope

3. Technical Support Contact: Robert Schneider, robert@pe.utexas.edu, +1 512 471 3225.

4. Project Abstract:

   Accurate and efficient reservoir simulation is of critical importance to the economic well-being of the United States. Future oil and gas production in this country is dependent upon improved oil recovery, reservoir simulation and reservoir characterization technologies. Intense computer simulation is essential for effective reservoir management. Parallel reservoir simulators have the potential to solve larger and more realistic problems than previously possible. Our research involves development of new physical and chemical models, accurate numerical methods, new reservoir description techniques and their implementation in parallel environments. In particular, we investigate porting of reservoir simulators to a distributed memory cluster of Intel processors. By performing this task, the feasibility of using a cluster of Intel processors for solving large-scale reservoir simulations will be assessed.

Back to Top

Real-Time Process Control
Status Report (January, 1998)

1. Principal Investigator: T. F. Edgar

2. Associated Investigators: Prof. J. Qin, Dept. of Chemical Engineering

3. Technical Support Contact: Tyler Soderstrom, 471-1046

4. Project Abstract:

   This project deals with the application of PC's for real-time modeling and control of chemical processes. The goal of the research is to demonstrate the use of Intel PC's in a real-time, computationally demanding environment where advanced optimization and control algorithms are employed. The experimental applications under study include a batch distillation column, a reactive distillation column (currently under construction) and a rapid thermal chemical vapor deposition system. In addition, we are using PC's off-line to carry out optimization and estimation with chemical dynamic simulators such as Dynaplus from Aspentech.

5. Equipment Status: So far we have received one dual Pentium PC and one single Pentium on 1/27/98.

6. Research Progress.

   The PC's are being used in an ongoing research program that involves 12 Ph.D students. In the Intel Project graduate students will be actively working with the Pentiums as part of their research, which is funded by a number of private companies, including Aspentech, Simulation Sciences, and Union Carbide. In addition, we have secured a major donation from Fisher-Rosemont Systems of their latest computer control software/hardware system, which is called Delta V (software just released in late 1997).

   
   

Back to Top

Subsurface Modeling
Status Report (May, 1998)

1. Principal Investigator: Mary F. Wheeler

2. Project Abstract:

   The Texas Institute for Computational and Applied Mathematics (TICAM) is a fully vertically integrated research environment spanning mathematical modeling, finite difference/volume/element methods, parallel solution methods, and visualization. The range of expertise represented by the faculty and research staff at TICAM presents and ideal opportunity to advance research in interactive steering for large scale parallel applications.

   The parallel projects and applications within TICAM currently include:

   • Subsurface Flow & Chemistry (IPARS, PIERS, PARSSIM)
   • Surface Flow (ParWQM, pADCIRC, UTBEST, UTPROJ)
   • Visualization (VisualEyes)
   • Computational Fluid Dynamics
   • Composite Materials
   • Electromagnetic Scattering
   • Data Structures (SDDA, DAGH)

   Under the Intel grant, the Center for Subsurface Modeling, a center under the auspices of TICAM, has been responsible for assembling a "Beowulf" class parallel host using 16 processor Pentium II nodes with dual ports of a 100Mbit network used by the operating system (Linux). The system is configured through a 100Mbit Intel data switch which is connected to the TICAM network.

3. Equipment Status:

   The system is up and running and the code PARSSIM is being tested on the system. In particular, the Beowulf system has been benchmarked as running 10% faster in parallel and compilation approximately an order of magnitude faster than on the TICAM SP2.

   The Beowulf system will be relocated on May 18 to temporary quarters for 18 months. We anticipate the system to be operation by June 1.

4. Research Progress:

   The Center for Subsurface Modeling has leveraged this system nodes and has obtained a grant of $255,000 from the NSF Center for Research on Parallel Computation to increase the size of the system.

Back to Top

Wireless & High Performance Telecommunication Networks
Status Report (January 20, 1998)

1. Principal Investigator: Gustavo de Veciana

2. Associated Investigators:
   Takis Konstantopuolos, San-qi Li, Ed Powers, Guanghan Xu

3. Technical Support Contact:
   Gustavo de Veciana, Aimin Sang, Xun Su

4. Project Abstract:
   

   This project is intended to support the core of telecommunications research faculty within the Electrical and Computer Engineering Department. The equipment will support compute intensive research on physical layer (satellite,wireless) transmission issues, measurement-based channel and traffic modeling, and protocols for traffic management in high-speed networks.

5. Equipment Status:

   As of March 25 our 9 Intel machines have been delivered and have been distributed among five faculty in the telecommunications area.

6. Research Progress.

   Industrial Interactions ---

   Southwestern Bell Corp. / Technology Resources Inc.

   Related Funded Projects ---

   NSF Grant - Adaptive Resource Management for IP/ATM Hybrid Switching Systems (PI: San-qi Li)

   NSF Career Award - Analysis and Design of Hierarchical Source Routing and Embedded ATM Netowrks (PI: G. de Veciana)

   Soutwestern Bell Corp./ Technolgy Resources Inc - Traffic Control and Management (PIs: S.-Q. Li and G. de Veciana)

Back to Top

Distributed Simulation
April 9, 1998

1. Principle Investigators: Dr. Craig M. Chase and Dr. K. Suzanne Barber

2. Technical Support Contact:: Craig Chase

3. Project Abstract:

   This project involves the development and demonstration of high-performance simulation using clusters of Intel-based computers and high-speed networks. The major areas of exploration are low-latency communication models, object-oriented simulation and distributed object environments. The fundamental precept underlying this study is that clusters of high-performance workstations can substantially replace or outperform special purpose parallel computers provided appropriate software models and networking technology is employed. Leveraging additional support from Tandem Corporation and utilizing the Tandem ServerNet System-Area Networking technology, we are developing the necessary software infrastructure to confirm this hypothesis.

4. Equipment Status:

   Equipment	Qty	Received
   Dual CPU 300MHz Pentium II systems	4	Yes
   Quad CPU 200MHz Pentium Pro systems	6	Yes
   Desktop 300MHz Pentium II systems	2	Yes
   20 High Resolution Monitors	4	3 of 4
   100Mbps Switched Ethernet hubs	2	Yes

5. Research Progress:

   The six quad-processor Pentium Pro systems have been in extensive use since early January. These systems have been performing simulations in support of computer architecture research and have provided the basis for the following submitted papers:

   a. Brian Grayson and Craig Chase, ³Techniques for Low-Overhead and No-Overhead Communication Using Commodity Superscalar Processors,² submitted to 1998 IEEE conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS).

   b. Brian Grayson, Lizy John and Craig Chase, ³The Effects of Memory-Access Ordering on Multiple-Issue Uniprocessor Performance,² submitted to IEEE Transactions on Computers.

   We have connected the ServerNet networking equipment to the dual processor machines. This small cluster forms our experimental base, and we have completed our initial benchmarking of this platform. Performance is currently severely limited by the inefficiency of the TCP/IP protocol implementation over ServerNet. We anticipate finding a software solution to this performance bottleneck in the coming weeks (we hope to identify a mechanism that supports the Compaq/Intel/Microsoft Virtual Interface Architecture). Benchmark results are available on the project web page: http://www.ece.utexas.edu/projects/distsim/

Back to Top

1. Principal Investigator: Larry Leibrock

2. Associated Investigators: Patrick Jaillet, Andrew Whinston

3. Technical Support Contact: Joe Castro, Nicole Shaw

4. Project Abstract:

   This project includes three exemplar aspects of enterprise systems: deploying a commercially viable mission-critical application set such as computational finance; assessing and managing a computationally complex real-time distributed electronic commerce application; and providing the fundamental performance assessment, best practices and management expertise of the Intel/NT environment, all within a larger heterogeneous network. The interplay of IT practitioners, real commercial needs, and university research will result in the realistic assessments of the potential powers, promise, and possible limitations of commercial off-the-shelf Intel Architecture distributed systems.

5. Equipment Status:

   The Texas Business School has received twelve quad processor servers that are being used in a variety of ongoing projects. The machines are running NT 4.0, NT 5.0, MSSQL 6.5, and IIS 3.0 and 4.0.

6. Texas Business School Projects URL:
   http://bevo3.bus.utexas.edu/cbacc/intel/
   

Computational Finance
Project PI: Patrick Jaillet

PROJECT DESCRIPTION

This project will test the capabilities of advanced Intel platforms for solving high-end computational finance problems in the area of pricing, portfolio optimization, and risk management. Leading-edge financial institutions routinely use advanced techniques from mathematics, engineering, and computer science to create, price, and manage risk for new instruments. High-end computers are now powerful enough to allow analyses and decisions based on real-time data feeds (tick-by-tick), as well as on huge real-time databases (data warehousing).

Specific computationally demanding problems to be addressed in this proposal will include (1) solutions to support activities linked to data warehousing and data mining; (2) fixed-income securities and derivative businesses; (3) valuation of securities with multiple sources of "noise," in multi-factor models (e.g., term structure models, FX models, and convertible securities); and (4) valuation of path-dependent instruments, especially mortgage-backed. This program is to develop and test distributed and parallel computing solution procedures for these most computationally demanding problems based on exclusive Intel Architecture - Microsoft NT system. The design techniques are aimed at providing a lower cost of ownership for the solution of the most computationally demanding finance problems.

TECHNICAL CHALLENGES

The complexity and fast resolution of this class of computational finance problems routinely require the use of very advanced techniques and algorithms derived from mathematics, engineering, and computer science. The need for high performance computing, high-levels of systems integrity, systems availability, and "fast-cycle" solutions becomes even more critical when "real-time" analyses and capital allocation decisions are needed. The computational finance decisional systems must be made based on very large streams of real-time data that is increasingly derived from electronic trading activity. A second challenge will be the integration of such procedures within hybrid networks allowing both NT machines and supercomputers/superservers (e.g. Cray).

IMPACT

Traditionally, "back-office’’ calculations have been developed on networks of UNIX workstations with access to fast superservers and supercomputers. Recently however, Window NT workstations have increasingly replaced the "front-office" infrastructure in financial markets (e.g., London, Wall Street), and have started to gain access to the "back-offices" as well. It is thus essential to evaluate and demonstrate the (1) "at the limit" performance of parallel and distributed computing Pentium-Windows NT architecture for such computationally intensive financial problems, and (2) the ease of their integration with traditional supercomputers. Proof of concept will increase the exposure of Intel Architecture - Microsoft NT system to the financial markets, and more generally into financial industries at large.

EQUIPMENT

Cluster NT Systems, Systems Lab with 32 dual processors workstations, fast communications between computers with 100 Mbit/s Fast Ethernet.

RESOURCES

This project will be centered in the Institute for Computational Finance (ICF) and the Texas Institute for Computational Mathematics (TICAM). The 35 affiliate faculty of ICF span three colleges (Business, Engineering, and Natural Sciences) and eight departments. Resources will also include facilities such as the EDS Financial Trading and Technology Center (FTTC), and the IC2 Institute.

BENEFITS TO INTEL

As discussed above, this project is to evaluate and demonstrate the at the limit performance of parallel and distributed computing Pentium-Windows NT architecture for computationally intensive finance problems. Proof of concept will increase the exposure of Intel Architecture - Microsoft NT system into the financial markets, and more generally into financial industries at large.

Back to Top

Electronic Commerce Networks
Project PI: Andrew Whinston

PROJECT DESCRIPTION

Creating and operating a digital economy is fast becoming a reality. It is expected that several hundred billion dollars per year of electronic transactions will take place by the start of the next century. The project will develop mechanisms for management of the network infrastructure that will enable effective commerce to take place and facilitate innovative applications exploiting the infrastructure. The Center for Research in Electronic Commerce has conducted pioneering activities in this area including the development of resource management mechanisms using priority pricing of network resources and estimation of response times from available data. While the ideas are based on concepts from economic theory the computational methods exploit an understanding of the transient behavior of public networks. The project seeks to extend the mechanisms previously demonstrated in simulation testing to real prototype networks. These mechanisms are characterized by their incentive compatibility and the decentralized resource management in real time. Techniques have been developed to estimate performance parameters over the network, and to update prices at each node of the network in real time based on such available information.

TECHNICAL CHALLENGE

The project addresses the larger issues of operationalizing any proposed mechanism. The decentralized real time price formulation rules we employed has been designed to be computationally manageable. The challenge in a prototype network is to enable running complex applications over the network in an efficient manner, and with reliable performance guarantees. The enhanced computing power, the predictable performance in real time, and the distributed nature of applications will be crucial determinants in realizing electronic commerce applications, for instance, in solving a supply chain management problem. Visions of electronic-commerce-transacted networked organizations have been hindered by the excessive congestion over networks and their unreliability. By charging prices that reflect the traffic at each node our mechanism manages congestion and provides predictable performance guarantees. Since a real time network does not support theoretical equilibria proposed by economists, our mechanisms are developed to handle non-equilibrium behavior.

IMPACT

The work will demonstrate the feasibility of managing resources over a computing network in real time, in an efficient manner, to achieve reliable performance levels. Further, the mechanisms used will be shown to be scaleable to larger systems, with no computational overheads. This will directly demonstrate that distributed computing models for complex tasks and networked execution of critical business processes can both be feasibly achieved over public networks. A computationally complex application (such as a Supply Chain Management solution) can be executed at much lesser cost over such a managed network of Intel systems.

EQUIPMENT

A network of Pentium processor based computers to execute distributed application will be displayed, while the usage is monitored and priced by software implementations of the pricing formulate. As a first prototype network, a cluster of 12 workstations and 2 Quad servers are to be used.

RESOURCES

This project is to be carried out at the Center for Electronic Commerce. The research will be an interdisciplinary initiative, with faculty from the departments of Management Science and Information Systems, Economics and Computer Science collaborating. Previous and ongoing work on developing the theoretical framework and testing it out by simulations has been supported by grants from NSF and the Advanced Technology Program of the State of Texas.

BENEFITS TO INTEL

An enterprise-wide application such as the Supply Chain Management problem is computationally complex, distributed in nature, and requires real time solutions, typically handled by high-end compute servers. Deployment of efficient network resource management mechanism over the connected PCs to solve such a problem is proposed. This will demonstrate that properly managed, a simple network of Intel PCs can pack the computing power and reliability of high end servers and will be capable of enabling Electronic Commerce with a broad reach.

RELATED WEB SITES

Virtual reality check

Back to Top

Enterprise Systems Management
Project PI: Larry Leibrock

The goal of Enterprise Systems Management is to provide information-technology based systems that are highly functional, reliable, of high quality, trustworthy and developed within the enterprise planned resource constraints. The network is heterogeneous and multivendor with more and more applications that are central to the operation of an enterprise being deployed in a multi-tier architecture that separates the business logic, the data, and the presentation onto multiple systems. This project has three primary goals. First, it will provide expertise to other projects in assessing the performance characteristics, capabilities and limitations of their computationally demanding tasks in an Intel/NT environment. Secondly, this project seeks to develop better enterprise security models and policies necessary to protect data in distributed heterogeneous systems architectures, And thirdly, the project seeks to evaluate model development in the context of MicrosoftÌs Internet Security Framework, NT based digital certificate key servers and the recently announced Intel ÏSecure CardÓ specification as replacement for current password models.

TECHNICAL CHALLENGES

Intel/NT systems are increasingly being utilized for mission-critical applications such as SAP, electronic commerce needs and trusted commercial requirements. Research on behalf of Intel can assess present security frameworks and provide for creation of best practices for digital key servers and ÏSecure CardÓ specifications in high-performance computational environments. Penetration testing (breaching security) is essential. Neither the industry nor research community have conducted methodologically based penetration testing for inter-operating computer environments (e.g. NT and UNIX) especially in business systems environments The boundaries between differing systems offers considerable risks to high-integrity requirements. In addition, this project will provide expertise and best practices for assessing the performance of computationally demanding applications in order to provide necessary assistance to other projects.

IMPACT

Corporations are increasingly using Intel/NT computational clusters and high availability systems to support mission-critical distributed environments such as BAAN, SAP and Oracle Financials. Typically, these applications are two and three tiered client server applications. The boundaries among the tiers of applications, networks and operating systems provide a considerable level of risk where security can be penetrated and critical data destroyed, stolen or altered.

EQUIPMENT

Server facilities and research labs are available for housing and supporting the equipment and technical support staff is already in place. The Business School, ACITS and TICAM will share use of the servers and desktop systems. This equipment will also support the Computational Finance project and other TICAM projects. The equipment includes 12 Quad servers, two sets of 22 dual graphic workstations, 4 Dual desktop workstations and a print server.

RESOURCES

Several faculty and doctoral students are already working in the area of enterprise systems. Other faculty have interests in applications of new technology and have published research papers and given talks at national conferences as well. Business School, ACITS and TICAM researchers have extensive experience in the development of both measurement tools and high performance requirements. The current research interests lie in migrating projects to a high-performance client-server environment in a distributed environment.

BENEFITS TO INTEL

Development of better algorithms, systems management policies, performance analysis, security testing and best-practices risk models would be of benefit to Intel and its enterprise customers as commercial mission critical applications begin to be deployed that require a scaleable, reliable, economically manageable, and secure multi-vendor distributed environment.

Back to Top