Table of Contents
-
New to supercomputing? Click
here.
-
KEYNOTE:
Irene
M. Qualters,
National Science Foundation
-
PLENARY:
Addison
Snell,
Intersect360 Research
-
PLENARY:
Henry Neeman,
University of Oklahoma
-
PLENARY:
Platinum Sponsor Speaker:
Stephen Wheat,
Intel Corp
-
Erik Alred,
University of Oklahoma
-
Daniel Andresen,
Kansas State University
-
Workalemahu M.
Berhanu,
University of Oklahoma
-
Manikanthan
Bhavaraju,
University of Oklahoma
-
Keith Brewster,
University of Oklahoma
-
Dana Brunson,
Oklahoma State University
-
Nick Davis,
University of Oklahoma - Tulsa
-
James Ferguson,
National Institute for Computational
Sciences
-
Karl Frinkle,
Southeastern Oklahoma State University
-
Zane Gray,
University of Oklahoma
-
Saeed Iqbal,
Dell
-
Robert Kern,
Mellanox Technologies
-
Houssain Kettani,
Fort Hays State University
-
Evan Lemley,
University of Central Oklahoma
-
Greg Monaco,
Great Plains Network
-
Mike Morris,
Southeastern Oklahoma State University
-
Don Schulte,
Cray Inc.
-
Chris Stackpole,
Federal Reserve Bank of Kansas City
-
Mickey Stewart,
Arista Networks
-
Bin Wang,
University of Oklahoma
PLENARY
SPEAKERS
Division Director
Advanced
Cyberinfrastructure Division
(ACI)
Directorate
for Computer & Information
Science & Engineering
(CISE)
National
Science Foundation
Topic:
"Collaborative Cyberinfrastructure to Explore
21st Century Research and Education Frontiers"
Slides:
PowerPoint
PDF
Abstract
Ubiquity in mobile devices,
social networks,
sensors,
advanced computing and instruments
have created
a complex data-rich environment
ripe for
new scientific and engineering advances.
In this world of
computational and data-enabled
science and engineering,
a dynamic and innovative yet cohesive
cyberinfrastructure of
technologies,
services,
and
people,
is fundamental to
all aspects of the discovery process.
This talk will focus on
National Science Foundation's
vision, strategy and support of
collaborative cyberinfrastructure.
Biography
Irene M. Qualters
is currently
Division Director of
Advanced
Cyberinfrastructure
(ACI)
at the
National
Science Foundation
(NSF).
ACI is responsible for
programs with
a total annual budget in FY2013 of
over $200 million.
These programs support the
acquisition,
development,
and
provisioning
of
state-of-the-art
cyberinfrastructure resources,
tools,
and
services
essential to the conduct of
21st century science and
engineering research and education.
ACI is also responsible for
the NSF-wide
vision,
strategy,
planning
and coordination
for research cyberinfrastructure.
She joined NSF as
a Program Director
in December 2009,
participating in
multidisciplinary,
interagency
and
international activities
as well as overseeing
several major computational projects
within the division's portfolio,
including the
Blue
Waters
project
at the
National
Center for Supercomputing Applications
(NCSA)
at the
University
of Illinois at Urbana-Champaign
(UIUC)
and the
Stampede
project at
Texas
Advanced Computing Center
(TACC)
at the
University
of Texas at Austin.
Irene has a Master's degree in
Computer Science.
Prior to beginning her NSF responsibilities,
she had a distinguished 30-year career
in industry,
with executive leadership positions for
research and development organizations
within the
technology sector.
During her twenty years at
Cray
Research,
in increasingly larger leadership roles,
she participated in
the development of
the first commercially successful
vectorizing compiler,
the first
multiprocessor version of
Unix,
and
Cray's landmark massively parallel computer,
the
T3E.
Subsequently,
for six years,
as Vice President,
she led the
Research Information Systems for
Merck
Research Labs
(MRL).
She is expert in
parallel computer system architectures
and in
a wide variety of software
from
scientific applications
to compilers
to file systems and operating systems.
Assistant Vice President/Research
Strategy Advisor
Information
Technology
Director
OU
Supercomputing Center for Education
& Research (OSCER)
Information
Technology
Associate Professor
College
of Engineering
Adjunct Associate Professor
School
of Computer Science
University
of Oklahoma
Topic:
"OSCER State of the Center Address"
Slides:
PowerPoint
PDF
Talk Abstract
The
OU
Supercomputing Center for
Education & Research
(OSCER)
celebrated its 13th anniversary
on August 31 2014.
In this report,
we examine
what OSCER is,
what OSCER does,
what OSCER has accomplished
in its 13 years,
and where OSCER is going.
Biography
Dr.
Henry Neeman
is the
Director of the
OU
Supercomputing Center for Education &
Research,
Assistant Vice President
Information Techology
–
Research Strategy Advisor,
Associate Professor in the
College
of Engineering
and
Adjunct Associate Professor in the
School
of Computer Science
at the
University of
Oklahoma.
He received his BS in computer science
and his BA in statistics
with a minor in mathematics
from the
State
University of New York at Buffalo
in 1987,
his MS in CS from the
University of
Illinois at Urbana-Champaign
in 1990
and his PhD in CS from UIUC in 1996.
Prior to coming to OU,
Dr. Neeman was a postdoctoral research
associate at the
National
Center for Supercomputing Applications
at UIUC,
and before that served as
a graduate research assistant
both at NCSA
and at the
Center for
Supercomputing Research &
Development.
In addition to his own teaching and research,
Dr. Neeman collaborates with
dozens of research groups,
applying High Performance Computing techniques
in fields such as
numerical weather prediction,
bioinformatics and genomics,
data mining,
high energy physics,
astronomy,
nanotechnology,
petroleum reservoir management,
river basin modeling
and engineering optimization.
He serves as an ad hoc advisor
to student researchers
in many of these fields.
Dr. Neeman's research interests include
high performance computing,
scientific computing,
parallel and distributed computing
and
computer science education.
Chief Executive Officer
Intersect360
Research
Topic:
"Software Evolution in HPC"
Slides:
PowerPoint
PDF
Abstract
Drawing from the latest research studies,
Addison Snell of
Intersect360 Research
will discuss the evolution of middleware and
application software in HPC.
With multi-core,
many-core,
and cloud all
altering the computational landscape,
we have seen significant shifts in
software models,
particularly in commercial HPC markets,
where open source
is becoming more prevalent.
This presentation will include excerpts from a
wide range of research studies,
including data on installations and budgets,
forecast models,
a deep-dive study on Big Data trends,
and a
soon-to-be-released study on
the benefits to U.S. industry of
government investment in supercomputing,
published in conjunction with the
U.S.
Council on Competitiveness.
Biography
Addison
Snell
is the Chief Executive Officer of
Intersect360
Research
and a veteran of the
High Performance Computing industry.
He launched the company in 2007 as
Tabor Research,
a division of
Tabor
Communications,
and served as that company's
Vice President/General Manager
until he and his partner,
Christopher
Willard, Ph.D.,
acquired Tabor Research in 2009.
During his tenure,
Addison has established Intersect360 Research
as a premier source of market information,
analysis,
and consulting.
He was named one of
2010's
"People to Watch"
by
HPCwire.
Addison was previously
an HPC industry analyst for
IDC,
where he was well-known
among industry stakeholders.
Prior to IDC,
he gained recognition
as a marketing leader and spokesperson for
SGI's
supercomputing products and strategy.
Addison holds a master's degree from the
Kellogg
School of Management
at
Northwestern
University
and a bachelor's degree from the
University
of Pennsylvania.
Director, HPC Solutions,
Platform Collaboration and Systems Division
Intel
Topic:
"HPC's Cadence – Onwards and Upwards"
Slides:
PDF
Talk Abstract
The HPC market segment's
insatiable demand for performance
appears to continue unabated.
Fueled by competition,
as well as national and economic security,
the adoption and expansion of
the application of HPC
continues to drive the need for
innovation of the tools themselves.
At the same time,
HPC platform complexity is increasing
and the cadence of
refinement of designs
and
new product introduction
continues at a challenging rate.
In this talk,
I review the environmental conditions
as they have continued to change.
Then I address Intel's key strategies
to corral myriad technology advances
into leadership HPC solutions for
rapid and confident deployments
while maintaining applicability to
the spectrum of HPC solutions deployments,
whether large or more modest.
Giving a sense of what is coming,
this talk is meant to inspire
readiness and broader innovation.
Biography
Dr. Stephen Wheat is
the Director of HPC Solutions
at
Intel.
In that role,
he is responsible for
the development of Intel's HPC
solutions business
strategy
and
the pursuit of that strategy through
platform architecture,
eco-system development
and
collaborations.
Dr. Wheat was previously
General Manager for Intel HPC;
in that role,
he
influenced
the deployment of several Top10 systems,
even
more
Top500
HPC systems,
and pursued the democratization of HPC
through his Missing Middle initiative.
He possesses a depth of knowledge and
breadth of experience
that gives him
a unique perspective in understanding
large scale HPC deployments.
Also at Intel,
he was
the Advanced Development manager
for the Storage Components Division,
the manager of
the RAID Products Development group,
the manager of
the Workstation
software and validation groups,
and manager of
the Supercomputing Systems Division (SSD)
operating systems software group.
At SSD,
he was
a Product Line Architect
and was
the systems software architect for
the
ASCI
Red
system.
Before joining Intel in 1995,
Dr. Wheat worked at
Sandia
National Laboratories,
where he performed
leading research in
distributed systems software.
In that role,
he created and led the
SUNMOS
and
PUMA/Cougar
programs.
A
1994
Gordon
Bell Prize
winner,
Dr. Wheat
has also been awarded Intel's prestigious
Achievement Award
and was a top pick for
the
HPCwire
People to Watch List
in
2006
and again in
2011.
He has a patent in
Dynamic Load Balancing in HPC systems.
Dr. Wheat holds a Ph.D. in
Computer Science
and has several publications on
the subjects of
load balancing,
inter-process communication,
and
parallel I/O in large-scale HPC systems.
On a personal note,
Wheat is also
a commercial multi-engine pilot
and
a FAA certified multi-engine, instrument
flight instructor.
BREAKOUT
SPEAKERS
Graduate Student
Department
of Chemistry & Biochemistry
University
of Oklahoma
Topic:
"Application of
Computational Biophysical Techniques to
Amyloid Systems"
Slides:
available after the Symposium
Talk Abstract
Amyloid fibrils
are mis-folded proteins
that have been shown to be associated with
many diseases,
including life-changing illnesses such as
Alzheimer's and diabetes.
In several cases,
the previously suspected
vectors of cytotoxicity,
large fibril formations,
have been shown to be
less toxic than
the smaller, oligomeric states.
The transition between
oligomers and fibril states
involves structural changes in
the individual protein chains,
creating the necessity for data relating to
the structure of these chains in
the amyloid aggregates.
Computational techniques
employing biophysics and
molecular dynamics simulations
allow chemists and computer scientists
to probe molecules
on scales inaccessible
in a normal laboratory environment,
that of very short time ranges
and single aggregate experiments.
In this talk,
I will discuss
the applications of
these techniques towards
the understanding of
physical and molecular factors
that explain amyloid fibrils structures.
Biography
Erik Alred
performed his undergraduate studies at the
University
of Houston
Department
of Chemistry.
While there,
he also worked as
a part time researcher at
MD
Anderson Cancer Center.
After graduating with a B.S. in Chemistry,
he went to work in
the chemical industry
as a research technician in
Albermarle
Catalysts
Analytical and Research Development Department.
He is currently working in
Dr.
Ulrich Hansmann's group
as a graduate student.
His current research interests include
all atom molecular dynamic simulations,
probing of structural properties of amyloids,
docking studies of small molecules,
method development for
novel computational techniques,
mechanisms of protein folding,
and computational study of
the protein aggregation process.
Associate Professor
Department of
Computing & Information Sciences
Kansas State
University
Director
Institute for Computational Research
Topic:
"Minimum Time, Maximum Effect:
Introducing Parallel Computing in CS0
and STEM Outreach Activities Using Scratch"
Slides:
PDF
Abstract
This talk discusses
our experiences and outcomes
using
Scratch
to teach parallel computing concepts to
students just learning about computer science.
We presented versions of this material to
middle school and high school girls
during a summer camp
and then to
undergraduate university students
enrolled in
an introductory computer science course.
Using the Scratch development environment,
students were able to
build, modify and observe
the changes in
the performance of applications
that utilize
multi-threaded, concurrent operations.
This includes some scenarios that
involve more advanced topics,
such as race conditions.
We first used this exercise
in Summer 2013 with
a group of middle school girls
as part of a summer STEM camp at
Kansas State University.
After our lesson,
22 of 41 students surveyed
showed an interest in
a job using high performance computing
to solve problems,
and 27 of 37 said
they were capable of learning
to write computer programs.
Biography
Daniel
Andresen, Ph.D.
is an associate professor of
Computing
& Information Sciences
at
Kansas
State University
and Director of the
Institute for Computational Research.
His research includes
embedded and distributed computing,
biomedical systems,
and high performance scientific computing.
Dr. Andresen coordinates the activities of
the K-State research computing cluster,
Beocat,
and advises the
local
chapter
of the
Association
for Computing Machinery
(ACM).
He is a
National
Science Foundation
CAREER
award winner,
and has been granted research funding from
the NSF,
the
Defense
Advanced Research Projects Agency
(DARPA),
and industry.
He is a member of
the
Association
for Computing Machinery,
the
IEEE
Computer Society,
the
Electronic
Frontier Foundation,
and
the
American
Society for Engineering Education.
Postdoctoral Research Associate
Department
of Chemistry & Biochemistry
University
of Oklahoma
Topic:
"Inter-species Cross-seeding:
Stability and Assembly of
Rat-Human Amylin Aggregates"
Slides:
PDF
Talk Abstract
Diseases such as
type 2 diabetes,
Alzheimer's
and
Parkinson's
share as a common feature
the accumulation of
mis-folded disease-specific protein aggregates
into fibrillar structures,
or plaques.
These fibrils may either be
toxic by themselves,
or act as reservoirs for
smaller cytotoxic oligomers.
This suggests
to investigate molecules
as potential therapeutics
that either reduce fibril formation
or increase fibril stability.
One example is rat amylin,
which can inhibit aggregation of human amylin,
a hallmark of type 2 diabetes.
We use molecular dynamics
to compare the stability of
various preformed aggregates,
built out of either human amylin,
rat amylin,
or
mixtures of both.
We considered
two types of fibril-like oligomers:
a single-layer in-register conformation,
and
a double-layer conformation
in which the first U-shaped layer
consists of rat amylin
and the second layer of human amylin.
Our results explain
the weak amyloid-inhibiting properties of
rat amylin
and suggest that membrane leakage
due to pore formation
is responsible for
the toxicity of rat amylin
observed in a recent experiment.
Together,
our results put in question
the use of rat amylin
or the similar FDA approved drug pramlintide
as an inhibitor of human amylin aggregation.
They also point to
mixed human-rat amylin fibril-like oligomers
as possible model-systems for
studies of amyloid formation
that involve cross-species transmission.
Biography
Workalemahu M. Berhanu
graduated
in 2011
from the
University
of Central Florida
with a PhD
in Chemistry.
Since January 2012,
he has been working in
Prof.
Ulrich Hansmann's
group in the
Department
of Chemistry & Biochemistry
at the
University
of Oklahoma
as postdoctoral research associate.
His research interest
has been targeted towards
understanding the mechanism of
protein folding,
protein mis-folding,
protein aggregation,
cross seeding of aggregates,
toxicity mechanism
and their inhibition
based on bimolecular simulations.
Postdoctoral Research Associate
Department
of Chemistry & Biochemistry
University
of Oklahoma
Topic:
"Structural Analysis of a Non-Amyloid Protein
and an Amyloid Fibril"
Slides:
available after the Symposium
Talk Abstract
The most common form of systematic amyloidosis
is associated with the light chain amyloidosis
and is related to the overproduction of
the free monoclonal light chains plasma cells
in the bone marrow.
Due to various physiochemical conditions,
such as
change in pH,
salt concentration,
temperature,
rare mutations,
etc.,
the light chain proteins undergo unfolding
and form amyloid fibril or aggregates.
Six different residues
were individually mutated
to explore the effect of
single point mutation on
a non-amyloid protein.
An amyloid protein was taken
and was mutated to
a non-amyloid protein
and vice-versa,
to analyze
the structural difference between them.
The root mean square fluctuation of
the residues 40-60 in
R61N, D82I, and A84T mutant systems
were higher than
a wild type non-amyloid protein
and may initiate the unfolding process.
Further,
an amyloid fibril model was considered
and various residues were altered
to build a non-amyloid fibril model.
For both the systems
the
Cα- Cα
distance between
various residues involved in
the monomer-monomer and dimer-dimer
interaction region
were measured.
The residues in the dimer-dimer region
were close than
the residues in
the monomer-monomer interaction phase.
Thus,
the dimer-dimer interaction
may play a major role in
the aggregation process.
Biography
Manikanthan Bhavaraju
attended graduate school at
Mississippi
State University
and received a PhD in
Chemistry
in 2013.
Since May 2014,
he has been working as a
Postdoctoral Research Associate
in
Dr.
Ulrich Hansmann's group
in the
Department
of Chemistry & Biochemistry
at
the
University
of Oklahoma.
During 2005-2007,
he worked as a Lecturer in the
Department
of Chemistry
at
Muffakham
Jah College of Engineering and Technology
in India.
His research interests include
atomistic molecular dynamics simulations,
docking studies,
theoretical calculations,
and studying
protein folding mechanisms and
aggregation processes.
Senior Research Scientist and
Associate Director
Center for
Analysis & Prediction of Storms
University of
Oklahoma
Topic:
"Interstate Data Moving and
the Last Block Problem:
Lessons Learned in CAPS Spring Experiment 2014"
Slides:
PowerPoint
PDF
Talk Abstract
In Spring of 2014,
for the first time,
CAPS endeavored to send complete
high resolution 4D
numerical weather prediction forecast files
from
the
National
Institute for Computational Sciences
(NICS)
in
Oak Ridge TN
to forecasters in the
Hazardous
Weather Testbed
in the
National
Weather Center
at the
University
of Oklahoma.
Having the complete files would allow
the use of 3D visualization tools
and exploration of the
forecast output not previously possible.
As weather forecasts,
the data were very perishable
and prompt reliable throughput was essential.
The issues around
moving very large datasets in real time
halfway across the country
are explored from the end-user perspective,
with some lessons learned and
some future solutions presented.
Biography
Keith Brewster is a Senior Research Scientist
at the
Center
for Analysis and Prediction of Storms
at the
University
of Oklahoma
and an Adjunct Associate
Professor in the
OU School of
Meteorology.
His research involves
data assimilation of
advanced observing systems
for high resolution
numerical weather analysis and prediction,
including data from
Doppler
radars,
satellites,
wind profilers,
aircraft
and
surface mesonet systems.
He earned an M.S. and Ph.D. in Meteorology
from the
University
of Oklahoma
and a B.S. from the
University
of Utah.
Director
High
Performance Computing Center
Adjunct Associate Professor
Department
of Computer Science
Oklahoma
State University
Topic:
"Advancing Research Computing
on Campus and Beyond"
Slides:
PDF
Abstract
The demand for computational resources,
support,
and expertise among the
academic community continues to grow
with each passing year
and can quickly
surpass
an institution's capacity
to meet researchers' needs.
The return on
investment
in these resources
can lag behind the initial investment
causing
challenges in developing
an institutional commitment for funding
centralized
highly available cyberinfrastructure.
In this session,
we will discuss the
approach taken at OSU
to develop these resources
utilizing external funding,
cooperation and collaboration with others,
and attention to specific needs
of the campus community.
Biography
Dana Brunson
is Director of the
High
Performance Computing Center
and is an adjunct associate professor in the
Computer
Science Department
at
Oklahoma
State University
(OSU).
She earned her Ph.D.
in Numerical Analysis
at the
University
of Texas at Austin
in 2005 and her M.S. and B.S. in Mathematics
from OSU.
program
Dana is co-lead of the
OneOklahoma
Cyberinfrastructure Initiative
(OneOCII)
and is a member of the
XSEDE
Campus
Champion
leadership team.
Assistant Professor of Research
Department of Medical Informatics
School
of Community Medicine
University
of Oklahoma - Tulsa
Topic:
"Rapid Calculation of
Medication Adherence
Using Parallel Computing with R and Python"
Slides:
PDF
Abstract
Medication non-adherence
is an issue affecting
the majority of Americans,
negatively impacting patient health outcomes
as well as being a burden on the U.S. economy.
Using electronic pharmacy claims data from
the
Oklahoma
Health Care Authority,
medication adherence is calculated for
a subset of patients
using the recommended
Proportion
of Days Covered
(PDC)
method.
Performance
across a number of languages and tools
is compared,
and the techniques employed
to parallelize the code discussed.
Reducing overall runtimes allows for
rapid
calculation and subsequent analysis
of the data.
Python
and its multiprocessing package
are shown to outperform
competing methods in execution,
and offer other advantages as well.
Biography
Dr. Nick Davis
is Assistant Professor of Research in
Medical Informatics at the
University of Oklahoma-Tulsa
School of Community Medicine.
He received his BS in
Computer
Science
with a minor in
Mathematics
from the
University
of Tulsa
in 2002,
his MS in
Computer Science
with a focus in
Information Security
from the
University of Tulsa
in 2004,
and his PhD in
Computer Science
from the
University of Tulsa
in 2012.
For his doctoral work at TU,
Dr. Davis performed research in bioinformatics,
focusing on
genomic analysis of
immune response data sets
and
analysis of
fMRI
brain imaging data
to identify regions of interest.
In addition to his academic experience,
Dr. Davis has accumulated
over a decade of industry experience in
a variety of technology roles,
such as software development and architecture,
network and system administration,
and information security,
including being a
Certified
Information Systems Security Professional
(CISSP).
Dr. Davis's research interests include
analysis of
electronic health record
and claims data,
data science algorithms and tools,
machine learning/statistical inference,
medication adherence,
integrative analysis of
heterogeneous biological data sets,
and
high performance computing.
Education, Outreach, and Training Director
National
Institute for Computational Sciences
University
of Tennessee Knoxville
Topic:
"XSEDE Campus Bridging Tools"
Slides:
PDF
Talk Abstract
While many researchers
are often able to easily get
allocations and accounts on national resources,
the overall impression for
a new user of large scale resources
is frequently,
"Now what?"
Campus Bridging attempts to make
national scale resources appear proximal to
the researcher
through a combination of
education and training,
data management
and
execution management services
that allow for users to access resources
as easily as if they were attached to
the user's own workstation
or other local resource.
The
XSEDE
Campus
Bridging
team
focuses on advocating
a common means of access to
national resources
(InCommon
authentication),
software packages for
campus system administrators,
gathering
requirements from
campus and local users on
software packages to be implemented by XSEDE,
and providing opportunities to interact
with new software for use in XSEDE.
Biography
Jim Ferguson
is the Director of
Education, Outreach & Training
for the
National
Institute for Computational Sciences
(NICS)
at the
University
of Tennessee Knoxville.
His responsibilities include
coordinating a wide range of
outreach and education related activities
associated with NICS,
as well as
varied responsibilities as
a level 3 manager for
Student Programs
in the
Education
& Outreach
area of the XSEDE project.
Jim has served on
many workshop and conference
organizing committees,
with current efforts including
the upcoming
SCxy
Conferences
and the
International
HPC Summer School
series.
Before joining NICS,
Jim's focus was programming for,
training,
and educating
users of
high performance computers and networks.
Jim's previous experience includes
positions at
Pratt
& Whitney Aircraft
and the
National
Center for Supercomputing Applications,
including significant roles in
NSF-funded projects
like the
National
Laboratory for Applied Network Research
and
Web100.
Jim is an alumnus of
Rose-Hulman
Institute of Technology.
Karl
Frinkle
Associate Professor
Department
of Mathematics
Southeastern
Oklahoma State University
Topic:
"Getting HPC into Regional University Curricula
with Few Resources"
(with Mike Morris)
Slides:
PowerPoint
PDF
Abstract
We showcase
our successful project of
injecting high performance computing (HPC)
into
the traditional computer science curriculum
at
Southeastern
Oklahoma State University.
We incorporated a three-semester sequence of
parallel programming courses,
with the third course focusing on
a research-level mathematical project
that was executed on OU's supercomputer.
Emphasis was placed on utilizing
Open MPI
and
CUDA
libraries along with
parallel algorithm analysis
and file I/O optimization.
We recruited
students with
varied CS backgrounds
for the program,
some with only CS-1 in
their portfolios.
These courses
are currently being followed with
a more hardware-oriented course this fall,
which will be profiled in the talk.
We will discuss our approach
to making this a successful sequence of courses
and divulge many of the tips and tricks
we have learned from our efforts.
Biography
Karl Frinkle
is an applied mathematician
who earned his PhD from the
University
of New Mexico.
He is deeply interested in
numerical simulations,
and most recently in parallel programming.
Karl joined
the SE Mathematics department in 2005,
and thoroughly enjoys teaching
parallel programming
courses
with
Mike Morris
through the CS department.
He also can be found teaching
physics courses in the Physics department.
IT Architect II:
Manager of Data and Telecommunications
Information
Technology
University of
Oklahoma
Topic:
"Synergistic Opportunities for
Computational Resource Providers and Local IT"
Slides:
PowerPoint
PDF
Talk Abstract
How do high-speed data paths
benefit research computing,
and should you care?
This talk explores
the means of engaging local
Information Technology resources
to build two-way, collaborative partnerships
that enable the achievement of research goals.
We will discuss examples of
how this relationship benefits local IT
through exposure to
cutting edge technologies
prior to introduction into
the campus or enterprise environment
—
a selling point in an affiliation that has
traditionally been more "take" than "give."
We will also discuss when circuit services,
Science DMZ's,
and Research and Education Networks
are appropriate enough
to warrant engagement with
local or regional IT resources.
Biography
Zane Gray is
the Data and Telecommunications Manager for
the
University
of Oklahoma's
Norman Campus.
He received his Bachelor of Science in
Chemical
Engineering
from OU in 1996.
Before returning to
OU as a network engineer,
he was employed as a
civilian with the
United
States Air Force,
and deployed computer networks
throughout the United States.
At OU,
he collaborates with a team of
highly talented network associates;
oversees the daily operations and maintenance
of the campus computer network;
charts the direction of
a network design
that supports tens of thousands of devices;
interacts with other
IT professionals,
faculty,
and
students;
and faces the day to day challenges of change
with a smile.
Principal Systems Engineer
Global Solutions Engineering - HPC
Dell
Topic:
"Performance and Power Implications of
Hardware Accelerators"
Slides
PDF
Abstract
The latest hardware accelerators
are powerful parallel processors.
These accelerators are capable of
providing quantum improvements in
performance and performance per watt
across a broad range of HPC applications.
However,
to fully realize these improvements
for a given application,
a balanced HPC cluster design is important.
This talk highlights
some of the major issues
encountered during such HPC cluster design.
We will also utilize
High
Performance Linpack
(HPL)
and other commonly used applications
to analyze performance and performance per watt
of compute nodes.
The general aim is
to provide information
and introduce some tools
to enable accelerator enabled
HPC cluster designs.
Biography
Saeed Iqbal
is a principal systems engineer
in the global HPC solutions engineering group
at
Dell.
Currently,
he is leading
hardware accelerator
performance characterization efforts.
He is also responsible for
the HPC advisor,
an online tool available at
Dell.com/hpc.
Customers can use
the HPC advisor
to design solutions
according to their requirements.
Previously,
he led the Beowulf clustering software,
Grid computing
and
virtualization solutions
advisor projects.
His interests include
performance modeling and analysis of
parallel and distributed architectures,
power-efficient system design
application accelerators,
application specific architectures,
signal and image processing,
artificial intelligence,
scheduling and load balancing algorithms.
He holds a Ph.D. in
parallel and distributed computing from
the
University
of Texas at Austin.
Regional Sales Manager - South Central
Mellanox
Technologies
Topic:
"The Future of Interconnect"
Slides:
available after the Symposium
Talk Abstract
Mellanox
InfiniBand technology
is the foundation for
scalable and performance demanding
computing infrastructures.
Delivering more than 100Gb/s throughput,
sub 700ns application to application latency
and message rates of
137 million messages per second
has already placed
Connect-IB
as the world leading interconnect solution.
We will discuss
the latest interconnect advancements
that maximize
application performance and scalability
and focus on
MPI 3.0
readiness with
Core-Direct
technology.
We will cover
best practices for
increasing CPU efficiency
by offloading CPU tasks to the network,
discuss
configurable hierarchical collectives
(HCOL)
and also cover
the new transport service named
Dynamically Connected Transport
service
(DCT).
DCT uniqueness is with separating
the interconnect resources from
cluster size
and ensuring
the highest performance capabilities
for unlimited application processes.
We will review
the superior capabilities of
Mellanox interconnect solutions
that are available today,
as well as into future generations.
Biography
Robert Kern
is a Regional Sales Manager at
Mellanox
Technologies.
After graduating in 1978 with a
Bachelor of Science degree in Mathematics,
his professional career began with
Datapoint
Corporation
in San Antonio.
He moved to California in 1991
and spent several years in the
integrated circuit photomask business.
A career transition to
SynOptics
Communications
in 1988
culminated in a promotion and move to Dallas
in 1990.
For the past 24 years, he has been with
a handful of manufacturers
and provided business solutions inclusive of
switching/routing,
firewalls,
application delivery controllers,
wireless arrays,
and cloud optimization.
He joined Mellanox in January 2013.
In his spare time,
he enjoys
restoring,
modifying,
and subsequently pushing
the performance limits on
old Pontiacs.
Professor and Director
Computer
Science and
Information Systems Engineering
Fort
Hays State University
Topic:
"Information Systems Engineering at
Fort Hays State University"
Slides:
PowerPoint
PDF
Talk Abstract
The speed of my new
Dell
XPS
laptop is
9 GigaFLOPs
(billions of floating point operations
per second,
such as multiplications).
In other words and in theory,
my laptop could perform
9,000,000,000 multiplications per second!
New desktops
would be three
times faster.
The fastest supercomputer is
Tianhe-2
in China.
It has over three
million processor cores
and consumes almost
18 MW (Mega or million Watts).
It is like 180,000 light
bulbs of 100W on
at the same time.
Tianhe's speed is
34 PetaFLOPs
(quadrillions of floating point operations
per second),
which is almost 4,000,000
times faster than my new laptop.
So,
one second on Tianhe is equivalent to
4,000,000 sec or 1,100 hours or 46 days
on my laptop.
Also one hour on Tianhe is equivalent to
450 years on my laptop!
These fast computers allow humans
to solve problems that were impossible to solve
a few years before,
including
weather (earth and space) forecasting,
gene permutations,
hurricane tracking,
asteroid/comet tracking,
spying,
etc.
However,
such humongous machines present
huge complexity in
operation,
maintenance,
protection,
etc.
This remains an active area of research for
continuous improvement in speed and efficiency,
and provides an example of the need for
information systems engineers.
Information is knowledge derived from data.
The latter are recorded facts and figures
that include
collection,
storage,
management,
retrieval,
analysis,
etc.
Thus,
Information Systems
is the collection of
hardware,
software,
data,
people
and
procedures
that work together
to produce quality information.
To this end,
Information Systems Engineering
is the effective use or
"engineering" of
Information Systems
for a better understanding of data
to optimize decision making.
Therefore,
ISE is an emerging engineering discipline
consisting of
a multidisciplinary area of studies
that involves
software,
digital storage and retrieval,
networks,
human computer interaction,
information security,
digital design
and
electronic media.
Students from the ISE Program
are prepared to work with
complex information systems
and have the skills and knowledge necessary
to advance steadily in their careers.
Biography
Dr. Houssain Kettani
is professor and founding director of
the
Bachelor
of Science in
Information Systems Engineering Program
at
Fort
Hays State University
in Hays KS,
which is one of two programs in
the United States
and one among a handful worldwide.
Professor
Department
of Engineering & Physics
University of
Central Oklahoma
Topic:
"Enabling Computational Research and Education
at the University of Central Oklahoma"
Slides:
PDF
Talk Abstract
Coming soon
Biography
Evan Lemley
received his BA in Physics from
Hendrix
College
and MS and Ph.D
in Engineering (Mechanical) from the
University
of Arkansas.
His thesis
work was focused on modeling and simulation of
various neutron detectors.
Post graduation Evan worked for
the engineering consulting firm
Black &
Veatch
in a group responsible for
modeling coal power plants with
custom written software.
In August 1998,
Evan became an Assistant Professor in the
Department
of Engineering and Physics
(formerly Physics)
at the
University
of Central Oklahoma,
and has been there since,
teaching
mechanical engineering,
physics,
and
engineering computation
courses.
Early research at UCO was focused on
neutron transport in materials.
More recently,
Evan has been involved in simulation of
flow in microtubes and microjunctions
and
simulation of flow in porous networks.
Director for Research &
Cyberinfrastructure Initiatives
Great
Plains Network
Topic:
"Enhancing Networking Expertise
Across the Great Plains"
Slides:
PowerPoint
PDF
Abstract
Major scientific breakthroughs are accomplished by teams of
scientists working together,
remotely,
and accessing distributed datasets
and advanced tools via the United States and global research networks.
As
new cyberinfrastructure technologies emerge to meet the growing needs of the
science community,
campus personnel at U.S. institutions must learn and
adopt the new technologies for U.S. scientists to remain competitive.
This
need extends to smaller colleges and universities where cutting edge
expertise may be less readily available.
For this project,
the Great Plains Network is partnering with experts to
improve knowledge,
expertise and research cyberinfrastructure at campuses in
Kansas,
Missouri,
Nebraska,
Oklahoma and South Dakota.
Partners are KanREN,
MOREnet,
OneNet,
Reed Network,
Network Nebraska,
Internet2,
ESnet,
University of Chicago and University of Oklahoma.
Project objectives include benchmarking and periodically reassessing
progress of campuses in the region toward implementing existing and emerging
cyberinfrastructure technologies,
developing and delivering an online
curriculum to improve expertise,
creating opportunities for campus
participants to learn from leaders in advanced cyberinfrastructure and
network technologies,
providing outreach to campuses and building a
community of support and encouragement for the implementation of new
technologies.
The project measures progress by growth in the number of active campus
participants across this project's lifespan,
growth in the number of unique
campuses participating in the project,
participant mastery of advanced
technology concepts covered in training,
and an increase in number of
campuses implementing these specific technologies.
This project promises to directly benefit scientists in the fields of
physics,
bioinformatics,
climate modeling,
and weather forecasting.
Biography
Dr.
Greg Monaco
has held several positions with the
Great
Plains Network
since August 2000, when he joined GPN.
He began as Research Collaboration Coordinator,
and then was promoted to
Director for Research and Education,
followed by Executive Director
for several years.
He is currently the
Director for Research and
Cyberinfrastructure Initiatives.
Assistant Professor
Department
of Chemistry, Computer and Physical
Sciences
Southeastern
Oklahoma State U
Topic:
"Getting HPC into Regional University Curricula
with Few Resources"
(with Karl Frinkle)
Slides:
PowerPoint
PDF
Abstract
We showcase
our successful project of
injecting high performance computing (HPC)
into
the traditional computer science curriculum
at
Southeastern
Oklahoma State University.
We incorporated a three-semester sequence of
parallel programming courses,
with the third course focusing on
a research-level mathematical project
that was executed on OU's supercomputer.
Emphasis was placed on utilizing
Open MPI
and
CUDA
libraries along with
parallel algorithm analysis
and file I/O optimization.
We recruited
students with
varied CS backgrounds
for the program,
some with only CS-1 in
their portfolios.
These courses
are currently being followed with
a more hardware-oriented course this fall,
which will be profiled in the talk.
We will discuss our approach
to making this a successful sequence of courses
and divulge many of the tips and tricks
we have learned from our efforts.
Biography
Mike Morris' degrees are in math,
but he has always said
he wound up on the business end of a computer.
He taught Computer Science (CS)
in the early 80s
after working as
an Operations Research Analyst for
Conoco
in Ponca City OK.
Mike left teaching and spent 15 years
doing various things in the CS industry
before returning to
Southeastern Oklahoma State
to once again teach CS,
where he remains today.
Senior Account Manager
Cray
Inc.
Topic:
"Cray Solutions for
Higher Education and Research"
Slides:
available after the Symposium
Talk Abstract
Cray has always been known for
its high-end, high-performance compute systems.
It has a strong presence at
the top national academic computation centers
in the world.
This talk will highlight
some of the pioneering science
done on these systems.
The focus of the talk,
however,
will be on some of
the less well known products
that
Cray brings to the academic research.
This includes
a standard cluster offering
that combines high quality standard components
with the software and services
that
Cray is known for;
and a full range of storage products.
The talk will cover
Cray's end-to-end solutions
that are optimized for
both compute and data intensive workload.
Biography
Don Schulte
is a 30 year veteran of
the computing technology industry,
primarily in the areas of
enterprise Unix and research systems.
After graduating with
a Computer Science degree,
for 18 years Don implemented
large aerospace and defense systems in
his field-based Systems Engineer role
for computer hardware and software companies.
Most recently,
for last 12 years,
Don has held business and sales positions for
Dell,
Cluster
File Systems,
Penguin
Computing,
DataDirect
Networks
and
Cray Inc.
Since 1999,
Don has been involved in
some of the largest
High Performance Computing systems
within the US research community.
Linux Cluster Administrator
Research Automation
Federal
Reserve Bank of Kansas City
Topic:
"My Journey to Shift
the Mindset of Interactive Users to
the Power of Batch Computing"
Slides:
PDF
Talk Abstract
Many of our users
are very serial in their thought patterns
and prefer
to see their computations interactively.
They don't quite understand
how to take advantage of
the parallel nature of a cluster
and are hesitant to trust batch systems
with their work.
The journey that I and my coworkers are on
is one to bridge those gaps.
We seek to simplify
their interaction with the cluster
and to instruct them on
how to think and compute in parallel.
Biography
Chris Stackpole
has been working in the HPC field since 2003,
when he first fell in love with
Parallel Programming.
He has worked in academia,
industry,
and government
on clusters ranging from
cluster of workstations,
to clusters built out of spare parts,
to purpose built clusters.
He always enjoys
a good conversation about
what people are doing with their cluster.
Senior Systems Engineer
Arista
Networks
Topic:
"The Value and Future of Ethernet in HPC"
Slides:
available after the Symposium
Talk Abstract
In this session,
we will discuss how Ethernet continues to be
used as a high-speed interconnect for
most of the commercial HPC clusters
in use today,
and why we shouldn't count it out
when looking to build
even the largest of cluster computers.
The roadmap of Ethernet is promising.
The arrival of
PCIe
4.0
will enable 100 Gbps connections to the server.
Ethernet will offer
25 Gbps and 50 Gbps connections based on
25 Gbps lanes.
100 Gbps Ethernet has been shipping in mass
for over a year,
significantly driving down cost,
and there are plans for 400 Gbps Ethernet.
Ethernet is leading the innovation for
high-speed interconnect technology
compared to other transport solutions.
Innovations such as
Remote
Data Memory Access
over Converged Ethernet
(RoCE),
iWARP,
and support for kernel bypass drivers
make Ethernet comparable,
and there are many advantages that
Ethernet offers over
other high-speed interconnect technologies.
Biography
Mickey Stewart has
more than 20 years experience in
computing and network technologies.
He works as a systems engineer for
Arista Networks,
specializing in
Data Center and High Performance Computing
architectures
using
the most modern and advanced
network operating system,
Arista
EOS.
Previously he spent time at
Cisco
Systems,
Chesapeake,
Paranet,
Anixter,
Compucom
and
Microsolutions.
He has held various
systems engineering,
solutions architecture
and
business roles.
Mickey has expertise in servers,
routing and switching,
unified communications,
network and information security,
storage and optical networking.
Mickey holds/has held
many industry certifications such as
CCIE,
CISSP,
CCDP,
CNE
and
CNX.
Assistant Professor
School
of Chemical, Biological &
Materials Engineering
University
of Oklahoma
Topic:
"Computational Simulations of
Electronic and Optical Properties of
Nanomaterials"
Slides:
PDF
Talk Abstract
Tunable electronic and optical properties of
nanomaterials
are desirable for
their applications in
electronic and optical devices,
molecular sensors,
and
catalysis.
In this talk,
I will discuss
two case studies of
computational simulations of
two-dimensional (2D) materials.
In the first example,
I will talk about
defects and impurities
and their effects on
the transport properties of graphene.
Second,
I will focus on
optical properties of
2D transition metal dichalcogenides.
In particular,
I will talk about
the large exciton binding energy
and
tunable band structure
and
optical spectrum of
MoS2 under tensile strain.
Based on these two examples,
I would like to demonstrate
how atomic-level computational simulations
can be helpful for tailoring
the electronic and optical properties of
nanomaterials.
Biography
Bin Wang is
an assistant professor in
the
School
of Chemical, Biological &
Materials Engineering
at the
University
of Oklahoma.
He received his Ph.D. in
Chemistry
from the
École
Normale Supérieure de Lyon
(France)
in 2010,
supported
by a
Marie
Curie Fellowship.
During his PhD,
he visited the
University
of Munich
(Germany)
for almost one year.
In 2009,
he was awarded a
Young Scientist Prize
for the best student paper at the
10th
International Conference on
Atomically Controlled Surfaces,
Interfaces and Nanostructures
(Spain).
Before
coming to OU this fall,
he was a postdoctoral research associate in
the
Department
of Physics and Astronomy
at
Vanderbilt
University.
His research is focused on
the computational simulations of materials
and
their applications in
energy storage and conversion.