Test Software

Website of Daniel W. Barnette

Exceptional Service in

the National Interest

Personal Information

Name: Dr. Daniel W. Barnette

Work address:

MS 1319

Sandia National Laboratories

Albuquerque, NM 87185-1319

Work phone/fax: 505-***-****

E-mail: *******@******.***

Professional Experience

1977-present: Sandia National Laboratories

Current title: Senior Member Technical Staff

Current Assignment: Scalable Computer Architecture 1422

Previous Assignment: Parallel Computational Sciences Department;

Aerodynamics Division

Education

June, 1984 Ph.D., Dept. of Aeronautics and Astronautics, Stanford University,

Stanford, California Dissertation title: "Numerical Investigations into the Parabolized

Navier-Stokes Equations"

June, 1977 M.S., Dept. of Aerospace Engineering, University of Texas, Austin, Texas Thesis title: "The Effects of Massive Blowby on the Initial

Trajectory of a Free-flight Rocket"

June, 1976 B.S., Dept. of Aerospace Engineering, University of Texas, Austin, Texas Under-graduate project and report: "Experimental Investigation of

Supersonic Flow Past a Wedge-Cylinder Configuration"

May, 1973 Associates Degree, Texarkana Jr. College, Texarkana, Texas

Technical Areas of Expertise at Sandia National Labs

2009

Project leader for Performance Modeling and Analysis Team.

2006 - 2008

Worked mostly with Performance Modeling and Analysis Team. The Application

Performance Modeling and Analysis Team's (PMAT) charter is to obtain and

analyze the runtime performance of Sandia applications and benchmarking

codes on current high performance computing (HPC) platforms. We strive to

aid software development teams, hardware acquisition teams, management,

co-workers, and colleagues in understanding and applying methodologies for

obtaining the highest possible performance.

Began studying Python programming language to use on problems associated

with developing, running, and storing results from benchmarking codes for

Sandia's HPC platforms. Began developing a user interface for compiling and

running codes as well as an interface to MySQL databases for storing

results. User interface is called PYLOT.

2003-2005

My work

with ASCI Red Supercomputer over this FY has involved:

Ongoing

improvement of evaluation test codes including:

Debugging test

code as needed

Improving run

scripts for eval tests

Documenting

significant changes re: eval tasks/procedures in web pages

Being assigned the

task to rewrite a major eval-test run script called Munops/IO-Munops,

which involved changing the script that ran previously only as root

to:

as root, setup

the run, partition the test machine, stop services, etc.;

as user, run

the test submission; and

as root, stop

the run, remove partitions, restart services, and/or reboot.

The

purpose of above assignment was to allow Computer Associates personnel

in San Diego to run these tests as user and without the need for having

root privileges. This task was completed and several eval tests have

been run with the new scripts.

Ongoing

maintenance, updates, and page creation for extensive eval test web

sight at Sandia

Participating in

weekly teleconferences with the ASCI Red team

Documenting eval

test results, including

a summary

report that documents each test suite s passes and failures

placing

relevant eval notes and documents on the eval web site

Archiving eval

test results on Sandia s SMSS mass storage system

Writing test plan

procedures for Janus eval tests

Tracking certain

test runs to document sub-test failure patterns

Working closely

with the Tflops OS team to resolve each eval test failure, if

failures occur

Updating the eval

CVS directory when changes are made to any eval test suite

Working closely

with Computer Associates personnel up until contract end date of Dec

31 2003; work involved running eval tests to help CA search for OS

problems

Being solely

responsible for building new Tflops OS versions and installing OS

patches on eval machines for testing (task taken over from Sue Kelly

when she moved to Red Storm full time)

Being solely

responsible for running extensive eval tests on all new OS versions,

taking approximately 1 week to complete while running tests on 4

mini-Tflops machines, most simultaneously

Being solely

responsible for running eval test on Janus as final assurance a new

OS version is ready for production

Being assigned the

task of installing and testing new PGI compilers for Tflops;

compilers are due to be delivered to Sandia soon

2000-2003 With Sue Kelly and others, ongoing evaluation of operating system

software integrity of Intel's Teraflops Supercomputer at Sandia; maintain

and improve evaluation software suite (500MBytes of source code); improve

Intel's evaluation web pages; develop and maintain Sandia's version of

evaluation web pages; also involved in developing test and evaluation

software for the Cray Red Storm Supercomputer to be delivered in 2004.

1998-2000 With Brian Jones and Erich Bender, Lockheed Tactical

Aircraft Systems, Ft. Worth, on Sandia's Shared Vision program, ported

Lockheed's production Computational Fluid Dynamics code, written in

Fortran using MPI, to Sandia's parallel compute clusters; obtained scaling

and speed-up data; helped test and debug operating system software on

parallel compute clusters using this code; report in progress.

With Rich Pryor, modified his genetic algorithms computer code,

written in C, for use on Sandia's parallel compute cluster; with John

Feddema, wrote controlling program to implement the resulting algorithm on Sandia's RATLER all-terrain

mobile robotic vehicles; field tested robotic vehicles with genetic

algorithm controlling program; developed and wrote visualization software

using ;

documented in SAND2000-2846.

With George Reis, retired, revised and improved Sandia's internal

ballistics code ONEGUN; report in progress.

1996-1998 With Curt Ober, developed techniques for applying parallel overset grid methods to

climate and aerospace applications; documented in SAND98-0701. Awarded

patent

#6,519,553.

With LDRD funding, worked on developing a theory for coupling

computational flow solvers with grid generation techniques so that the grid is generated

simultaneously with the flow solver with very little user interaction required;

documented in .

Awarded

patent

#6,356,860.

1993-1996 Oceanography and aerospace applications and code development for massively

parallel and cluster parallel computers; part of work documented in .

1993 Applied and developed porous media flow codes for applications on parallel

computers. With Julie Swisshelm, worked in area of ocean basin modeling and oil spill modeling, with goal of

implementing current/updated models on parallel computers; work documented

in .

1991-1993 Continued work in area of computing the flow around spinning bodies,

including finned reentry vehicles and wind turbines.

Performed aerodynamic analyses on

missile configurations, including an earth penetrator and the Multiple Launch Rocket

System.

1990-1991 Worked with EAGLE surface and grid generation codes; wrote EAGLET, an

EAGLE post-processing code that allows the user to extensively manipulate and prepare

grids for other codes.

Worked with F3D Navier-Stokes code using multiblock grids generated

by EAGLE and post-processed by the PEGSUS block-to-block interpolation code; and used

TECPLOT for graphics.

Worked on spinning finned/non-finned reentry vehicles, flat-nose

cone-cylinders, and two-dimensional rigid parachute shapes using Navier-Stokes

flow field codes and grid generation.

Worked with Al Hodapp on cables in tow; documented in .

1988-1990 Began study of the different sublayer models in the Parabolized

Navier-Stokes code used at Sandia. Implemented an improved sublayer model and global

iteration in the code.

Utilized thin-layer Navier-Stokes codes to generate solutions over

flight vehicles with large separation regions, both as continuation and stand-alone codes.

Analyzed several grid generation codes obtained from various sources to determine which

codes will be of most use to Sandia-related projects.

1987 Documented improved version of a code which calculates the dynamic

response of vehicles subjected to blast waves. Derived analytical test cases to compare

with computational results for code verification. With John Yio and Robert

Isidoro,

implemented this version of the code on the Test Data Reduction Division's Data General

computer.

1986-1988 With Mary Walker, worked on improving the parabolized Navier-Stokes

(PNS) code

used at Sandia. This work focused on stability of smoothing terms, the turbulence model,

and the wall boundary conditions; documented in,,

and .

Also worked on improved six-degree-of-freedom code for

vehicles subjected to blast waves, as an extension of previous work.

1983-1985 Together with Glen Whiting and John Andersen, both from Sandia, was

involved in the preliminary design phase of two bombs; documented in

SAND85-0797.

With John Kraabel, Sandia

Livermore, have analyzed blast effects on the trajectories of reentry vehicles using

explosively-driven shock tubes, wrote a six-degree-of-freedom computer code to take

measured pressures from the vehicle subjected to the blast and calculate its

trajectory; documented in SAND report.

1981-1983 Attended Stanford University after working part-time on doctoral degree at

Sandia Labs. Together with Prof. Joseph Steger, arrived at an iterative method of

computing flow fields using the parabolized Navier-Stokes equations which render the

equations less susceptible to instabilities induced by refining the computational grid.

Wrote dissertation.

1981 Together with Dwayne Bennett of Sandia, performed a Monte Carlo dispersion

analysis for the B61 bomb. Analysis indicated that the bomb was well within the

requirements specified by the military; documented in SAND81-1896.

1979-80 Together with Charles Karnes of Sandia Labs, wrote a trajectory code for

analyzing the dynamics of non-rigid vehicles. Application for the code involved an 80-ft

seabed corer dropped from a height of 200 ft off the ocean floor; documented

in .

1977-1978 Together with Ron Greene, et al. from Sandia Labs, extensively analyzed

wind tunnel data on the ERB (Extended Range Bomb) and Tiger II missiles. Examined the

effects of free-floating rollerons on the rolling moments of Tiger II;

documented in SAND report.

Contact this candidate