Cyberinfrastructure News
To see past content, see the blog archive.
Calendar - May 2008
Learn about IU's supercomputers from the experts!
Monday, May 12, 11:00am-1:00pm - ICTC Main Lobby, IUPUIYou're invited to stop by and visit with the system administrators who run IU's supercomputers, Big Red and Quarry -- two of the most powerful supercomputers in the world.
You're welcome to ask questions, share ideas, or just chat in aninformal setting. This is an opportunity for you to learn how these centralized research computing resources can help support your work.
Research Technologies Round Table
Thursday, May 29, 12:30-1:30pmIUB: Radio-TV Building Room 180
IUPUI: ICTC Room 497
Live URL: mms://wms.indiana.edu/rt_round_table
Science Gateways
Marlon Pierce
After more than a decade of development, tools for Web-based access to computing resources and data archives are now very mature. This month's Roundtable will include discussion of these Science Gateways and the Grid middleware that they access, architecture and standards used by the science portal community, component-based Web portals, Web Services, and workflow (or service orchestration) tools. Also discussed will be Web 2.0 and Cloud Computing approaches to resource and data access, and these tools' eventual merger into Science Gateways and portals.
Please note the new Radio-TV Building location for IUB. Videoconferencing equipment will broadcast to the IUPUI location.
Posted on: 05/01/2008 | 0 comments
Don't forget to optimize!
When you submit a job to one of IU's supercomputers, there will be hundreds of jobs in front of yours and more lining up behind. Optimization ensures all these jobs will run as quickly as possible.
How can you simply and quickly optimize your program? Let the compiler help. By using three types of compiler switches - machine, usage, and optimizing - you can make your program fast and sleek without a lot of work. In this article, we'll look at optimizing switches only, with a C program compiled using Big Red's IBM compilers as an example.
Not all switches improve a program's performance. Some are neutral, some will help or hurt based on combinations, and some will slow a code down dramatically on their own. Safe bets include -Q to replace function calls with the function's source code, and -qunroll to unroll loops as if they were sequentially written source code. Those that should be used with care include -qhot for combining loops.
The -O switch is a set of macros for these types of smaller optimizations. This switch optionally includes a digit (e.g., -O3) which increases optimization risks as it increases. To offset this, IBM offers the -qstrict switch, which tells the compiler not to take risks if there is some doubt. A good first test of our sample C program might be:
xlc -q64 -O3 -Q -qunroll -qstrict -qarch=ppc970 -qtune=ppc970
-qenablevmx -qaltivec -o my_project my_project.c -lm
Every program is unique, but a good compiler can typically improve the performance of even a well-written program by 1/3. Reading your compiler's man page and experimenting with a few simple optimization switches may even cut your job's run time in half.
IU's High Performance Applications group may be able to help with compiling and optimizing your programs.
Posted on: 05/01/2008 | 0 comments
Old GPFS scratch files to be deleted
Beginning June 1, 2008, files that haven't been accessed in more than 180 days will be automatically removed from IU's General Parallel File System (GPFS) scratch file system.
The GPFS scratch file system is accessible from Big Red, Quarry, and Libra.
Please be sure to back up important data stored on GPFS since it is a temporary scratch file system and is not intended for long-term storage.
For information on backing up data to the Massive Data Storage Service (MDSS), see the KB article "At IU, how do I use SFTP or SCP to access my MDSS account?."
Posted on: 05/01/2008 | 0 comments
New job type on Big Red
A recent update to Big Red's LoadLeveler resource manager includes a new job type: MPICH. LoadLeveler jobs of type MPICH, as opposed to "parallel", automatically generate host files for use with mpirun's "-machinefile" switch.
A job's host file can be referred to via the new $LOADL_HOSTFILE variable; use of the llmachinelist command and subsequent cleanup are no longer necessary!
In addition, the variable $LOADL_TOTAL_TASKS may be used with mpirun's "-np" switch. $LOADL_TOTAL_TASKS is calculated from the LoadLeveler job script keywords "node" and "tasks_per_node". As a result, changes to job geometry only need to be made with these keywords; the mpirun command's arguments will be modified automatically.
Finally, the MPICH job type provides support for some additional MPI application enhancements which we intend to implement in the coming weeks. We encourage all of our MPICH users to modify their job scripts accordingly.
Posted on: 05/01/2008 | 0 comments
IEEE 2008 eScience Conference Call for Participation
Organizing committees for the 2008 IEEE eScience Conference are now accepting papers and proposals for tutorials; posters, exhibits, and demos; workshops and special sessions on topics related to eScience, grid, and cloud computing. The conference is being hosted by Indiana University in partnership with Microsoft Research and will take place on December 7-12, 2008 at the University Place Conference Center in Indianapolis, Indiana.
For more information on topics of interest, submission guidelines and deadlines please visit the conference Web site.
Posted on: 05/01/2008 | 0 comments
Drug Discovery on Big Red
The continued rise in computational power is significantly enhancing medical research and promises to expedite the process of translating research in basic science to the clinic. High throughput computational screening as well as highly intensive biomolecular simulations are accelerating the process of drug discovery to a new level. Blessed with the nation's second largest School of Medicine, as well as a rich high performance computing cyberinfrastructure, such activity is a natural fit for IU.
Showcasing this is work by Dr. Samy Meroueh and his group in the IU School of Medicine's Department of Biochemistry and Molecular Biology, along with the Center for Computational Biology and Bioinformatics. He is using IU's Big Red supercomputer in research aimed at the discovery and design of small molecules that can be used for the treatment of various diseases.
Big Red allows Dr. Meroueh to multiply many times the rate at which he can identify potentially useful molecules, a process that involves screening millions of candidate compounds and computing how well the candidate molecule shuts down the disease causing agent. His work has led to the discovery of highly potent agents that block processes that tumors must undergo to metastasize. Dr. Meroueh is now working in collaboration with various faculty at the world- renowned Simon Cancer Center at the Indiana University School of Medicine to assess the effectiveness of these molecules in animal models.
For more information, please visit the Meroueh web page.
Posted on: 05/01/2008 | 0 comments
Bulk email from Big Red
If you run a large number of jobs through Big Red (or any other supercomputer), you should include the line
in your job scripts, unless you're absolutely positive your email server will accept a large number of emails. Read on to learn why.
Recently, a prominent research group stepped up testing of their system, running close to 9,000 jobs on Big Red each day. Each job sent a notification of the result to the group's Gmail account. Thousands of emails in a short time, forwarded to an external mail server, led to Big Red being classified as a spammer by SpamHaus. This meant that anyone at any organization using SpamHaus to create its blacklists couldn't receive email from Big Red - including Big Red users at IU!
IU sysadmins can't make sure every server out there will accept large numbers of emails. And since "Big Red" is not a real email server - it's an alias that aggregates the notification email from1024 nodes - we cannot use a standard bulk emailer setup to become a "legitimate" bulk emailer. If you run a lot of jobs, please help us to help you - turn off job notifications so you can still get the mail that you need.
Posted on: 04/04/2008 | 0 comments
Student in Open Systems Lab wins Microsoft Award
Congratulations to Joseph Cottam, of the Pervasive Technology Labs at Indiana University, on receiving the first ever Microsoft Award,from the International Network of Social Network Analysis (INSNA) for his paper "Extended Assortivity and the Structure in the Open Source Development Community." The award, given to authors of outstanding papers that address social relations aspects of softwaredevelopment, includes a $1,000 cash prize.
Cottam wrote the winning paper with Professor Andrew Lumsdaine, Director of the Open Systems Lab where Cottam works as a Graduate Research Assistant. The Cottam and Lumsdaine paper introduces a set of tools they call "Developmetrics" to investigate community formation and product development in the open source software community. Read the paper at http:// cs.indiana.edu/~jcottam/pubs/sunbelt2008.pdf
Posted on: 04/01/2008 | 0 comments
Big Red & Quarry queue changes
Big Red has a new SERIAL queue that includes all the nodes from theNORMAL and LONG queues. This will allow users to submit to the SERIAL queue single node jobs that are ideal for backfilling withouthaving to choose which queue will start a batch job sooner. An IEDC queue has also been added that will allow access to Big Red through the Indiana Economic Development Corporation in partnership with Purdue University.
Quarry also has a serial queue to serve the same purpose that includes all the nodes from the normal and long queues in addition to 5 dedicated nodes. A normal queue has also been added for jobs less than 7 days and includes 33 dedicated nodes in addition to the nodes from the long queue. In order to make these changes possible the long queue has been reduced to 42 nodes.
Please see the following Knowledge Base documents for more details:
Posted on: 04/01/2008 | 0 comments
The least you need to know updated!
The general introduction to Indiana University's Advanced Cyberinfrastructure is frequently updated, and has been recast in audio and audio/video format, for those who want to learn about computing, data storage, visualization and consulting while working on their aerobics (or who prefer a non-visual modality, for whatever reason).
These items are available from http:// rtinfo.uits.indiana.edu/documentation/.
Posted on: 04/01/2008 | 0 comments
Calendar - April 2008
Digital Library Brown Bag
Wednesday, April 2, 2008, 12:00-1:15pmWells Library E174
The Digital Library Program's Project Proposal Process
Stacy Kowalczyk and Michelle Dalmau
Learn about IU's supercomputers from the experts!
Thursday, April 10, 11:00am-1:00pmJordan Hall Atrium, IUB
You're invited to stop by and visit with the system administrators who run IU's supercomputers, Big Red and Quarry -- two of the most powerful supercomputers in the world.
You're welcome to ask questions, share ideas, or just chat in aninformal setting. This is an opportunity for you to learn how these centralized research computing resources can help support your work.
Digital Library Brown Bag
Wednesday, April 16,, 2008, 12:00-1:15pmWells Library E174
Sakaibrary Update: Integrating Library Resources with Sakai
Jon Dunn
Research Technologies Round Table
Thursday, April 24, 12:30-1:30pm
IMU Persimmon Room & ICTC 497
Why optimize my code? It works already!
John Samuel, Director of the Center for Statistical and Mathematical Computing, will discuss Star-P, a software package designed to extend and parallelize high-level languages, including
- a discussion of Star-P and how it extends Matlab,
- optimizing Matlab code by incorporating Star-P data types, and
- a review of Star-P availability on the IU campuses.
Posted on: 04/01/2008 | 0 comments
UITS provides support for award-winning IU musicologist
Congratulations to IU Jacobs School of Music musicologist Thomas Mathiesen, who in December 2007 took home his third Deems Taylor award from the American Society of Composers, Authors and Publishers for his work as an editor of the book Music and Ideas in the Sixteenth and Seventeenth Centuries.
Mathiesen serves as director for the IU Center for the History of Music Theory and Literature (CHMTL), a joint venture of the Jacobs School of Music and the Office of Research and the University Graduate School. With support from UITS Core Services, which provides hardware and server administration, Mathiesen and associate director Peter Slemon maintain the CHMTL Web site. The CHMTL site provides scholars from around the world with electronic access to doctoral dissertations in musicology and texts on music theory, aesthetics, history and literature, as well as historical music texts.
Posted on: 03/03/2008 | 0 comments
Fetal alcohol spectrum disorder data repository
For the past four years Indiana University researchers from the departments of Computer Science, Psychology, Anthropology, Anatomy and Cell Biology, and Medical and Molecular Genetics have been part of the Collaborative Initiative on Fetal Alcohol Spectrum Disorders (CIFASD), funded by the National Institute on Alcohol Abuse and Alcoholism. CIFASD, which includes researchers from 21 institutions in eight countries, was recently renewed for five years.
According to CIFASD Administrative Core Director Ed Riley of San Diego State University, "Fetal alcohol spectrum disorder is a major, worldwide public health issue and fetal alcohol syndrome is among the most common known causes of mental retardation in the western world. The work being done by the CIFASD is an international collaboration to study this problem using a multidisciplinary approach and should help define the spectrum of effects resulting from prenatal alcohol exposure. This includes providing better diagnostic protocols and an enhanced understanding of the changes in brain and behavior that occur following prenatal alcohol exposure, with the hopes of translating these findings into enhanced interventions."
CIFASD researchers are collecting a wide variety of data, including results from questionnaires, physical examinations, neuropsycho- logical test batteries, three dimensional facial images, brain images, and ultrasound movies. Data are collected at sites around the world using a variety of methods and then uploaded to the CIFASD Central Repository at Indiana University. CIFASD researchers can then use the Central Repository to join data from different locations and of different modalities. Members of the Research Technologies' Biomedical Applications group created the Central Repository and a variety of related software. A member of the Research Technologies' Advanced Visualization Laboratory developed the procedure used to capture three dimensional facial images.
The Central Repository for the CIFASD is hosted by the Indiana University Research Database Cluster, and makes use of the Indiana University Massive Data Storage System. The IU Research Database Cluster provides both an Oracle relational database management system and a Web application hosting environment used by the Central Repository. The Massive Data Storage System holds backups of the data in the Central Repository, providing highly resilient disaster recovery due to its geographically distributed nature.
More information is available on the Biomedical Applications group.
Posted on: 03/01/2008 | 0 comments
Indiana Academy students step into their virtual worlds at IU
Students from the Indiana Academy for Science, Mathematics, and Humanities may be a few years away from owning their own homes, but six juniors and seniors, with support from Indiana University's Advanced Visualization Lab (AVL), are able to step into a virtual kitchen of their own design.
The experience is part of an ongoing AVL project with an advanced computer class offered at the Muncie residential high school for gifted and talented students. Students from the program receive instruction from their computer science teacher, Susie Sechrest, and expert advice from IU's AVL staff, in creating a virtual environment of their own imagination.
Using programming knowledge in languages such as Java and C++, the students work in teams to create a variety of virtual environments, such as a mad scientist lab, an apartment, and a village. The students develop their projects in Virtual Reality Modeling Language (VRML) on a standard personal computer, and then interact with their creations in immersive 3D in the AVL=92s state-of the-art virtual reality theater.
Sechrest said her students have greatly benefited from their experience. "I could never relay the amount of information that can be absorbed through the use of the 3D glasses, the theater, and the big computer screen," she said. "This experience will stay with the students for a lifetime."
For more information on the Indiana Academy for Science, Mathematics, and Humanities or IU's Advanced Visualization Lab.
Posted on: 03/01/2008 | 0 comments
IU SPEA class invites you to take a survey
Staff from the UITS Research Technologies division have been working with a graduate level SPEA class this semester. The students are doing research on "green computing" and are studying efficiency and energy use in the IU data center. In an effort to learn more about high performance computing at IU and the people that use those resources, the students have put together a short survey which takes less than 5 minutes to complete. To request a copy please send email to hps-admin.
Posted on: 03/01/2008 | 0 comments
Calendar - March 2008
Digital Library Brown Bag
Wednesday, March 5, 2008, 12:00-1:15pmWells Library E174
The NEW IUScholarWorks: Repositories, Journals, and Scholarly
Publishing
Julie Bobay and Randall Floyd
This talk is co-sponsored by the Continuing Education Committee of the Bloomington Library Faculty Council.
CGB Roundtable: Marlon Pierce
Tuesday, March 18, 12:00-1:00pmMY 209
Tools for Web-based access to computing resources and data archives have been developed for over a decade and are now very mature. This talk will briefly review examples of these "gateways" and the associated grid middleware, and then delve into the architecture and standards used by the science portal community to develop reusable software components. Topics to be discussed include component-based Web portals, Web Services, and workflow (or service orchestration) tools.
Systems based on heavy-weight "enterprise" standards are being challenged by lighter-weight "Web 2.0" approaches that incorporate social networks, gadgets, content syndication, rich client interfaces (based on the resurgence of JavaScript), and REST-based services. The talk will review these and discuss the eventual merger of these technologies into science portals and gateways.
Digital Library Brown Bag
Wednesday, March 19, 2008, 12:00-1:15pmWells Library E174
The Evolution of Library Descriptive Practices
Jenn Riley
Research Technologies Round Table
Thursday, March 27, 12:30-1:30pm
IMU Persimmon Room & ICTC 497
Why optimize my code? It works already!
Ray Sheppard, Acting Manager of the High Performance Applications group, will discuss compilers and optimization. This talk will center on when it would be beneficial to spend time optimizing source code (and when it would not). Discussion of the time spent vs. gain realized will include case examples from users of Big Red. Simple tricks which may be used and some of the possible pitfalls to watch out for will be discussed as well.
Posted on: 03/01/2008 | 0 comments
TeraGrid '08 - Call extended
June 9-13, 2008 - Las Vegas, NV, USA
The Call for Participation for TeraGrid '08 has been extended, and all interested individuals and organizations are invited to participate. New deadlines are:
-
Paper abstracts (500 words maximum) - March 18, 2008
Full papers (7-10 pages) - April 1, 2008
Proposals for papers on original, innovative work will be received and reviewed in three tracks: Science, Technology, and Education. All papers will be peer-reviewed. The Call for Participation is available at: http://www.tacc.utexas.edu/tg08/
Posted on: 03/01/2008 | 0 comments
TeraGrid workshop at AAAS meeting
Technologists from Indiana University Information Technology Services will inform some of the nation's top research scientists about advanced cyberinfrastructure at the annual meeting of the American Association for the Advancement of Science (AAAS) being held next month in Boston. IU is sponsoring a tutorial and workshop titled "The TeraGrid: An essential tool for 21st century science," on Sunday, February 17, from 10:30-noon.
"The AAAS meeting is among the largest gatherings of the best scientific minds in the world," said Craig Stewart, Associate Dean for Research Technologies and chief operating officer for Pervasive Technology Labs at Indiana Unversity. "This workshop provides a tremendous opportunity to inform scientific researchers across a wide spectrum of disciplines about the TeraGrid, and to show how it can help accelerate innovation and enable new discoveries."
Workshop attendees will learn how the TeraGrid can support virtual organizations - distributed teams and communities of scientists that share common interests or needs for data and computing resources. Noted Stewart, "The NSF is promoting the concept of virtual organizations as a way of responding rapidly to today's scientific, medical, and security challenges. The TeraGrid's combination of computing power, storage capability, and science gateways offers an unparalleled mechanism for enabling virtual organizations to solve some of the most challenging problems facing scientists today.
This workshop will also show how the TeraGrid is becoming more accessible to all scientists - even those without computer science expertise - through Science Gateways, tools that make it easier to access and use supercomputers. Any scientist whose research is slowed or inhibited by limitations on computer power or storage capabilities will find this workshop and tutorial valuable.
For more information, visit the Research Technologies home page.
Posted on: 02/01/2008 | 0 comments
Calendar - February 2008
Lunch with a system administrator
Come and have lunch with an IU cluster admin. Mid-February, details in the Message of the Day.
Cell/Broadband Engine programming workshop
On February 4 and 5, the IBM Future Technology Solutions Center team will be conducting a workshop on Cell/Broadband Engine programming at IUPUI. Another workshop will be scheduled later this spring.
Digital Library Brown Bag
Wednesday, February 27, 2008, 12:00-1:15pmWells Library E174
Jon Dunn and Jenn Riley
This talk is co-sponsored by the Continuing Education Committee of the Bloomington Library Faculty Council.
Research Technologies Round Table
Thursday, February 28, 12:30-1:30pm
IMU Walnut Room & ICTC 497
George Turner
The power of today's computing technology is in its diverse and distributed nature. Tying together this vast cyberinfrastructure are tools such as Globus, GridFTP, and GSI-SSH. Research Technologies will offer a brief introduction into these grid tools, with a goal of stimulating discussion on how they can be used to increase research productivity.
Posted on: 02/01/2008 | 0 comments
Institute for Digital Arts and Humanities created
A newly established Indiana University institute is set to digitally redefine scholarship and creative activity in literature, music, dance, and many other arts and humanities fields.
The Institute for Digital Arts and Humanities (IDAH) will enable and expand digitally-based arts and humanities projects by bringing together scholars, artists, librarians and IT experts. The Institute draws on established strengths at the IU Bloomington campus in combining arts and humanities disciplines and information technology such as the Variations digital music library, the EVIA digital video archive of ethnographic music and dance, 3-D virtual reality work by IU artists and the IU Digital Library Program.
For more information, see the the press release or visit the IDAH home page.
Posted on: 02/01/2008 | 0 comments
TeraGrid Knowledge Base taking off
There were more than 19,000 accesses to the TeraGrid Knowledge Base during the month of December, which now has 291 entries. Here's one of the newest: "On Big Red or Quarry, why is my job sitting in the queue, and when will it run?" The answer involves the commands showq, checkjob, and showstart; it can be found at
http://www.teragrid.org/cgi-bin/kb.cgi?docid=awgw
or by searching the Knowledge Base for 'job queue' (for example).
Posted on: 02/01/2008 | 0 comments
Star-P available on Quarry
Star-P, from Interactive Supercomputing, Inc., is now available for user access on the Quarry cluster. Star-P is a client-server parallel computing platform that=92s been designed to work with Very High Level Language client applications such as Matlab or Python. Star-P makes parallel application development much more accessible than traditional programming models.
Vectorized Matlab code adapts well to Star-P's structures. Star-P supports parallelism by overloading core Matlab functions, so Matlab users should have little trouble incorporating Star-P. While not every Matlab Toolbox has explicit support for parallelism, seventeen commonly-used toolboxes do, including the neural network and signal processing toolboxes. A web-based tutorial on =20 using Star-P with Matlab is available.
Star-P's support for Python is also stable. We are interested in contacting active Python users for testing purposes. At present R support is in the development stage.
If you are interested in trying Star-P, the SoftEnv keyword for is +Starp. Once this key is added to your .soft file, you can start Star-P using =20= a command such as
starp -j'-l nodes=3D1:ppn=3D2' -p 2
If you need additional information on Star-P, please don't hesitate to contact the Stat/Math Center (statmath @ indiana.edu).
Posted on: 02/01/2008 | 0 comments
Special Announcements - January
[1] CGB Roundtable: Geoffrey Fox, January 22
[2] Special Seminar: Rick Rashid, January 24
[3] Cell Programming Workshop: February 4-5
CGB Roundtable
Tuesday, January 22, 12:00 PM
Location: Myers 209
Geoffrey Fox: Robust High Performance Optimization for Clustering, Multi-Dimensional Scaling and Mixture Models
Description:
We will first review the pros and cons of various approaches to nonlinear optimization in the presence of local minima, ill- conditioned matrices and ambiguous choice of appropriate number of degrees of freedom (over and under fitting). We define constraints on approaches from need to run well in parallel on systems of multicore CPU's.
We will present a uniform approach to data clustering and Gaussian mixture model ling that uses deterministic (not Monte Carlo) annealing to mitigate the local minima problem and naturally relates the appropriate number of parameters (clusters or mixture components) to the scale at which problem is examined. New clusters (mixtures) are introduced at phase transitions as the annealing temperature is lowered and second derivative matrix becomes singular. We contrast three ways of visualizing this structure in low (2) dimensions with Principal Component Analysis (PCA), Generative Topographic Mapping (GTM) and Multi-Dimensional Scaling (MDS), using annealing to regularize GTM and MDS.
Currently we have implemented, in a preliminary fashion, deterministic annealing clustering and GTM in a fashion that runs well on multicore systems. We have applied these techniques to Geographical Information Systems (clustering demographic data in 2D) and Cheminformatics in 1024 and lower dimensions. We would like other applications that can constrain and test these techniques.
More information: www.infomall.org/salsa
Special Seminar
Thursday, January 24, 4:00 PM
Location: Jordan Hall A100
Rick Rashid, Senior Vice President, Research, Microsoft: 10 years into the future
Abstract:
By looking at technologies in research labs today you can get insights into what opportunities technology will enable during the next 10 years. In this talk I will look at some exciting research technologies and their implications for the world of 2017.
Cell Programming Workshop
February 4th and 5th, 8:30 AM - 5:00 PM
Location: IUPUI, Informatics Complex, 535 W. Michigan, Indianapolis
Description:
On February 4 and 5, the IBM Future Technology Solutions Center team will be conducting a workshop on Cell/Broadband Engine programming.
This two-day Cell training will include a comprehensive technical overview of the Cell architecture, programming models, and software development environment. The training includes an introduction to many Cell programming concepts such as function off-loading, data streaming, software managed cache, module overlay, PPE and SPE instruction sets, and synchronous and asynchronous threading model.
In addition, use of the compiler to vectorize code, performance profiling, and code optimization will be covered. Participants should bring their own laptop. Detailed instructions will be made available to participants so that they can install the appropriate software for doing the hands-on exercise during the course. Laptops should have at least 8GB of free disk and at least 512KB RAM (1GB preferable).
Lunch will be provided.
To attend, please complete the following Class Registration Information and send it to xd2skco@us.ibm.com no later than January 23:
Name:
Title:
Company or University:
Office Phone:
Cell Phone:
email address:
Special Dietary Request:
Please Note: Registration is limited. You will receive confirmation no later than January 25. If you are not confirmed for this workshop you will be automatically registered for the next workshop which will be scheduled later this spring.
Class logistic information (directions, parking information,class room number) will be included with your class confirmation.
Posted on: 01/22/2008 | 0 comments
IU wins bandwidth competition at SC07
A team led by Indiana University was awarded first place in an international competition for leading-edge, high-bandwidth computing applications. The award was presented November 15 at SC07, the world's largest conference for high performance computing and communications. The team achieved over 91% of the theoretical peak performance in both directions, using production codes and production facilities, in particular the Data Capacitor. More information is available.
Posted on: 01/10/2008 | 0 comments
International Conference on e-Science and Grid Computing
The e-Science 2007 conference, sponsored by the Computer Society's Technical Committee for Scalable Computing (TCSC), is designed to bring together leading international and interdisciplinary research communities, developers, and users of e-Science applications and enabling IT technologies. The conference serves as a forum to present the latest research and product/tool developments, and highlight related activities from around the world. The conference will be held December 10-13, in Bangalore, India. More information is available.
Posted on: 01/10/2008 | 0 comments
Flybase expands to 12 genomes, funded $20M
Researchers at Indiana University Bloomington, Harvard University, Cambridge University (U.K.) and the University of California at Berkeley have been told by the National Institutes of Health that their request for approximately $20 million in continued funding for FlyBase has been approved. Biologists Thomas Kaufman and Kathy Matthews will oversee IUB's extensive contributions to the ongoing project. Kaufman was elected a fellow of the American Association for the Advancement of Science last month.
The most recent developments in the project -- sequencing and comparing 12 fruit fly genomes -- was the cover story of the November 8th issue of Nature, and has produced at least 40 papers, over a hundred new genes and several new regulatory pathways. Kaufman, who co-led the project, remarked "One of the things we've learned is that when you compare a lot of different but related genomes, you are more likely to see the genes that are buried in all that A-C-T-G mush." More information is available.
Posted on: 01/10/2008 | 0 comments
Massive Data Storage System use case
For the past two years Gary Pavlis of the IU Bloomington Department of Geological Sciences and School of Informatics has been involved in a collaborative project to study links between geologic structures and glacial erosion. The laboratory is in the Chugach- St Elias region of southeast Alaska, a unique place on Earth with very active mountain building in an area with the largest glaciers in North America.
Indiana is working with the University of Alaska on the seismology component of this project, monitoring earthquakes to help identify active faults and using modern imaging technologies to use global earthquakes to image the lower crust and upper mantle. The data are fed to the University of Alaska through spread-spectrum IP radios that provide links to VSAT terminals or sites with internet access.
The MDSS system at IU recently became a component of this project. Until then, only sections of the data were available at IU, due to the difficulty of transferring 500 Gbytes of data already acquired. With help from staff in the Research Storage group, an extremely clean solution to transfer data from the machine in Fairbanks to IU through the internet was developed. It provided the added benefit of a direct transfer to a long-term archive.
Posted on: 01/10/2008 | 0 comments
We're bloggin' now!
The IU Cyberinfrastructure Newsletter has a new incarnation, as a blog (with space for your comments!) and a new URL:
This will also be available as an RSS feed. The text-only version will continue to be provided via email. We hope that you enjoy the show!
Posted on: 01/10/2008 | 0 comments
Digital data toolkit funded by NSF
Beth Plale, associate professor in the Indiana University Department of Computer Science, has been awarded a National Science Foundation grant to develop a digital toolkit to help researchers more easily capture information about their scientific work.
"Because there is growing interest in storing scientific data to digital libraries, we are working with colleagues in the Digital Library Program at Indiana University to understand what provenance of scientific data is necessary for long-term preservation and use of an object," explained Plale. "In the past these multi-step, computational analysis tasks would require a script handwritten by a scientist, and annotation of the data would all be done by hand after-the-fact."
"Provenance of scientific data is an emerging research area, and one of importance not only to scholars, but to industry as well," added Plale, who directs the Center for Data and Search Informatics. "As the volume of scientific data from computational analysis grows into the petabyte range, it is important that information travel with the scientific data, wherever it eventually resides."
Additional information is available.
Posted on: 01/10/2008 | 0 comments
Calendar - January 2008
Lunch with a system administrator
Come and have lunch with an IU cluster admin. On January 14, from 11:00am until 1:00pm, IU cluster admins for BigRed and Quarry will be having lunch in the Maple Room of the IMU. Stop by and ask questions, express concerns, or just chat in an informal setting. Bring your lunch if you want - you can buy it in the IMU. Feel free to pass this on to anyone who might be interested.
Research Technologies Round Table
The Research Technologies Round Table is moving to the last Thursday of the month, to conflict less with other recurring events.
Thursday, January 31st, 12:30-1:30
ICTC 497 and IMU Sassafrass Room
Randy Heiland, manager of the High Performance Applications group [formerly HPC], will present an overview of their activities and welcome questions from the audience. The mission of the HPA group is to help promote scholarly research through the use of high performance computing and communication environments. The HPA group works closely with the High Performance Systems (HPS) group to fulfill this mission. A sampling of HPA activities include: user support for IU faculty, staff and students who want to get started using our supercomputers, longer term 1-on-1 consulting, support for our NSF TeraGrid users, benchmarking new or upgraded systems, and (in the future) developing services that hide the complexity of using high performance applications.
Posted on: 01/10/2008 | 0 comments
Programming support use case
Professor Charles Horowitz of the IU Physics Department and the Nuclear Theory Center carries out theoretical studies of dense nuclear matter, with the aim of understanding Type II supernovas, and the properties of neutron stars. Matter in these extreme astrophysical phenomena can reach almost unimaginable densities, beyond a trillion grams per cubic centimeter. At this density, atoms are completely ionized. At yet higher densities, the ions themselves lose their identities and turn into a soup of neutrons and protons.
Yet even at these high densities, matter can still be studied with a semi-classical model very similar to the kind used to study simple liquids. Simple liquids are modeled by assuming each atom exerts a central force on all the rest. The motion of each atom is found by simply integrating Newton's law, F=ma. In a molecular dynamics (MD) simulation, time is broken up into a series of small intervals. At the beginning of each time interval, all the forces are computed, and the corresponding accelerations are integrated to update the velocity and position of each particle.
The same type of thing can be done for the astrophysical situation. The most time consuming part of a molecular dynamics simulation is calculating all the forces. For a system of N particles, it takes on order of N squared calculations to get all the forces, and a physically realistic simulation may require anywhere from 10,000 to 100,000 particles. This means 100 million to 10 billion force calculations per time step. Moreover, tens of thousands to millions of time steps must be carried out to do a complete simulation, and many such simulations may be necessary to properly explore the physics.
Professor Horowitz's molecular dynamics simulations are a perfect fit for a special purpose computer called the MDGRAPE-2, of which IU has four. The MDGRAPE-2 performs the central force calculations in special hardware that does only that type of calculation, but does it as much as 50 to 60 times faster than a general purpose computer. With the help of IU's four MDGRAPE-2 machines, he has been able to carry out MD studies of supernova and neutron star matter that would have taken years on a serial machine. The High Performance Applications team has participated in this effort for several years.
Posted on: 01/10/2008 | 0 comments
Calendar - December 2007
There will be no Research Technologies Round Table in December.
Posted on: 01/10/2008 | 0 comments
System Administration
The maintenance window for all systems is the first Tuesday of
each month, 7am - 7pm EDT.
This includes the Massive Data Storage System!
Outage reports are available online for Big Red, Libra, and Quarry.
How To Contact Us And Get Help
If you have questions pertaining to IU's cyberinfrastructure, or you are encountering some difficulty, there are several ways to obtain help.
An introduction and overview titled "Indiana University's CyberInfrastructure: The least you need to know" has been updated and is available along with other introductory resources.
The IU Knowledge Base is an excellent source of help on how to do things.
If you have problems which the KB does not enable you to solve, questions about system outages, or if you just have a problem and you don't know who to contact, send email with the RT contact form.
This is an official publication of Indiana University produced by the Research Technologies division, University Information Technology Services, Indiana University.
Subscription is automatic when you receive an account on IU's advanced cyberinfrastructure. Please email comments, questions, and subscribe/unsubscribe requests, using the RT contact form.
Copyright 2007, The Trustees of Indiana University. Please credit UITS when referring to or using information in this publication.
