Entries filed under “Applied HPC”

Applications of HPC which are interesting because they enabled discoveries, represent new or uncommon domains for high end computation, or because they are, um, interesting.

GATech Creates Institute for HPC Research

The Georgia Institute of Technology Office of the Provost formally announced today the formation of the Georgia Tech Institute for Data and High Performance Computing (IDH).  HPC has become continually important to George Tech in order to further their research activities not directly related to the computational sciences.

As we look to high performance computing to drive advanced breakthroughs in science, health, energy and other industries, leveraging Georgia Tech’s strongest assets — world-class researchers in computing, experts across nearly every problem domain, and low barriers to collaboration — is what will set us apart,” said Dr. Mark Allen, senior vice provost for Research and Innovation at Georgia Tech. “The creation of the Institute for Data and High Performance Computing provides the organizational foundation to harness our strategic capabilities and attack the most challenging problems that face society today.”

The main goal of the IDH will be “to enhance Georgia Tech’s scientific contributions, reputation and impact, focusing on the exploitation of HPC technology coupled with the development of novel computational methods.”  The institute’s interim director will be Dr. Richard Fujimoto, Regents’ Professor and head of Computational Science & Engineering in the College of Computing.

Georgia Tech has made substantial infrastructure and personnel investments in high performance computing, and achieved many important successes, over the last five years,” said Dr. Fujimoto. “I fully anticipate that IDH will enable us to advance beyond prototypes to new levels of accomplishment in the high performance computing area.”

The IDH has quite the list of forward thinkers in computational science, including: Dr. David Bader, Professor in the College of Computing, and Dr. Ron Hutchins, Chief Technology Officer for Georgia Tech.  For more info, check out the new IDH website here.

Also posted in Computing Research | Leave a comment

IBM POWER7 system at Rice offers free cycles for medical research

Chron.com, the online presence of The Houston Chronicle, reported yesterday on a new super to be installed at Rice University. The system is one of the early POWER7-based systems shipped by IBM

IBM logoIBM has given Rice one of its first supercomputers with the company’s new POWER7 microprocessors, a $7.6 million award, on the condition that Rice open up access to the computer to all in the Texas Medical Center. For free.

This effort builds on earlier collaborations between TMC and Rice, and peaks out at about 18 TFLOPS

When combined, Rice’s ‘Blue BioU’ system is accordingly 18 times more powerful than Deep Blue. It doubles Rice’s existing supercomputer capacity.

Just one planned use? Genomics

Baylor College of Medicine is participating in a federal project to ascertain the genetics of 20,000 tumor samples from 20 different cancers….But Baylor doesn’t want to just store the information, it wants to cross reference each of these 100 DNA sequences with the originally sequenced version of human DNA to see which gene mutations might contribute to ovarian cancer.

So they’ll try to run the analysis on Rice’s new machine. Worley expects it will take hours, when it now takes weeks.

More in the article.

Also posted in Collaborations, HPTC, New Installations | Leave a comment

How to Build a Better Bobsled

Late night infomertial?  Not likely.  PCWorld has posted a series of very flashy photos courtesy of the simulations used in conjunction with the US Olympic bobsled team.  According to the photo stream sidebars, the US team has jumped on the HPC bandwagon to help them gain the vital fractions of a second in their chase for the gold.

This image [see the link below] shows some of the magic at work. Supercomputers and Exa PowerFlow software combine to improve the aerodynamic performance (shown by the green streamlines) of the sled by simulating the airflow over the vehicle as designers make each change. How effective has this approach been? Unbelievably so: The four-man Night Train team used the technology to design, hone, and build its amazingly fast bobsled, which won the 2009 World Championship last February in Lake Placid, New York–the first such title for a U.S. team in 50 years. Team Night Train hopes to transform that success into a gold medal in these Olympics when the bobsled events begin February 20.

If fluid dynamics is one of your big interests, check out the full photostream here.

Also posted in HPC | Leave a comment

Supercomputing apps migrate backwards

My pal Trey sent me an email yesterday with a pointer to an article in Computerworld about applications that are migrating backwards: going from supercomputers to desktops. This isn’t a new thing, but the advent of GPUs and CUDA has made it even more reasonable for non-heroic users to move to desktops if they can. There are good incentives for them to do this if they can still get their work done: no queues, no fighting with the big boys for allocation, and direct control over the resource.

Some of the interesting apps in the article

At Temple University, researchers have developed models that measure the effects of applying anesthesia on molecules within nerve cells. The models currently run on a supercomputer, but plans are underway to perform the calculations on an Nvidia GPU cluster with four nodes. This will both save money and give researchers more flexibility to conduct tests when they’re ready to do so (instead of having to wait for their scheduled time to use a supercomputer).

…Researchers at the Ohio Supercomputer Center (OSC) in Columbus, Ohio, have found that not every simulation requires a traditional supercomputer. Don Stredney, the director and interface lab research scientist for biomedical applications at OSC, found a limitation that’s common with supercomputers: Batch processes are static and run on a scheduled time frame. They cannot provide real-time interactions, so they can’t mimic a real surgical procedure. Desktop workstations that cost $6,000 to $10,000 allow his team to run simulations that show, in real-time, how a surgery changes a patient’s anatomy, he says.

…Injection-molding simulations are invaluable to car makers, Autodesk’s Martin says…Simulations “used to require a significant cluster-computing installation, but we are achieving the same level of power with current desktop computers,” says Martin, who says the desktop advances that played the biggest roles in making that possible were the move to multicore processing, the use of multiple GPUs and 64-bit throughput. Martin uses standard desktop computers that can be purchased at Wal-Mart, with the latest 3D-capable GPUs and Intel dual-core CPUs.

More interesting, complex applications in the article.

Leave a comment

Aalborg University Taps AccelerEyes Jacket

Researchers at Aalborg University in Denmark are now utilizing GPUs to advance their research in communications.  The Department of Electronics Systems and Technology Platforms focus on the theory, modeling, design, implementation, and test of RF systems and circuits, mainly within RF-CMOS, HW/SW co-design and design space exploration, advanced circuit and system theory, modeling and analysis as well as synthesis.  The research are using the AccelerEyes Jacket platform alongside MATLAB in order to accelerate development, irrespective of whether the target platform is pure CPU of CPU/GPU.

We often have few production runs and frequent model changes, which means that Jacket is an ideal solution for us,” says Prof. Larsen. He continues saying, “with Jacket we gain a lot of performance for a very modest investment in learning to take advantage of Jacket. We can’t afford to use a lot of time to develop C/FORTRAN programs – we rather let the computers together with Jacket do the hard computations. The speed-ups we have seen from Jacket compared to CPU solutions are very, very impressive.”

According to the release, the adoption of Jacket by Larsen’s team at AAU has resulted in a unique collaboration between AccelerEyes and AAU. Dr. Larsen provides continuous feedback to the engineering and development team at AccelerEyes to advance Jacket. Most recently Dr. Larsen has launched Torben’s Corner on the AccelerEyes Wiki to provide valuable information to the Jacket user community, globally.

For more info on the research and collaboration, read the full release here.

Also posted in GPUs, HPC Hardware, HPC Software | Leave a comment

Bull on Planet51

Earlier this month French HPC and IT company Bull announced that its servers handled the rendering for the 3D animated movie Planet 51

Planet 51The Planet 51 project was born in 2002. Since then, Ilion Animation Studios has created an entire parallel universe around Planet 51 using leading-edge technologies. The result is an incredible visual experience which cannot be compared with any other animated film. Over 350 people from more than 20 countries worldwide have worked on Planet 51 – including designers, developers, engineers and many other professionals – with a budget of some $70 million.

Bull logo…The Extreme Computing solution provided by Bull consisted of a cluster including computing nodes and a management node, connected to the storage system and other Ilion servers via an Ethernet Gigabit network.

Bull has proven expertise in Extreme Computing and in the implementation of Microsoft’s Windows HPC Server, a platform which is characterized by its high capacity for parallelism and its Job Scheduler system.

Within a month, the Bull supercomputer was ready to run, thanks to the involvement of a team of a dozen specialists from Bull France and Spain. The solution was clear proof of the high level of technological expertise of the team involved in the project.


Also posted in Enterprise HPC | Leave a comment

NTU Opens HPC Center, Installs IBM

The Nanyang Technological University [NTU] opened its High Performance Computing Center today.  The center’s charter being to assist the university’s ever growing research contingency, domestic and abroad.  The center also announced the installation of their first machine, an IBM iDataplex.  The 29Tflop machine will be used for research topics in developing future energy sources, studying global climate change, designing new materials, and understanding biological systems and the physics of complex socio-economic systems, among others.

Prior to installing the supercomputer, pockets of computing capabilities were located in Schools on campus which compete for space and financial resources. The establishment of the supercomputer brings under one roof a centralised large-scale computing facility to the 2,800-strong research community on campus. It will also pave the way for a wider range of complex multidisciplinary research endeavours and more opportunities for research interaction within as well as outside of NTU,” says NTU Provost Professor Bertil Andersson.

For more info on the new installation, read the full release here.

Also posted in Compute, HPC, HPC Hardware, New Installations | Leave a comment

Cray Forms Business Alliance With IRL Race Car Designer

Next generation IndyCar Series race car chassis designs are quickly coming to fruition.  The 2012 design year sneak peeks were leaked by race car company Swift Engineering.  Alongside the release was a side note on various new business partnerships that Swift has formed in the design and implementation process.  Oddly enough, Cray was highlighted as a major business partner.

Individually, each one of these partnerships is critical to Swift’s continuing commitment to motorsport; however, collectively and in conjunction with one another they will help us set new industry standards in innovative design, manufacturing and support,” Jan Wesley Refsdal [Swift's President] stated.

According to the source article, Cray’s machines will be utilized at the Swift research facility to augment CFD design processes.

Speed-to-market is critical in any business, but probably more so in racing as the green flag doesn’t wait for anyone,” Refsdal said. “Rapid development is just as much about the speed and quality of the design process as it is manufacturing. We are evolving our four-year exclusive certified composite repair relationship with Mark One Composites, Inc. to provide further manufacturing and inventory support directly to teams from its Indianapolis-based facility.”

Very cool stuff.  For more info, read the source article here.

Also posted in Compute, HPC, HPC Hardware | 1 Comment

IBM Builds Jeopardy Supercomputer

ibmI’ll take Deep Blue for 500 Alex.  Not quite.  IBM’s latest foray into human-like reasoning is beating humans at the ever-popular Jeopardy game show.  The question answered machine nicknamed “Watson” is already performing trial runs against humans that have previously appeared on actual Jeopardy episodes.  According to the article, Ken Jennings is not among the initial contestants.

Watson is “working its way up through the ranks,” says David Ferrucci, leader of the project team. “We win some, we lose some. Overall, we’re quite competitive but there’s a ways to go to play the top of the top.”

The games are taking place at the Watson Research Center in Yorktown Heights, NY where an actual Jeopardy stage has been setup for dry runs.  Alex Trebek is not present, but a real human is sitting in as a sample host.  Humans face off against “Watson”, who is located behind a glass wall.

But the Watson development team faces many challenges in creating a robotic Jeopardy champion. Without being connected to the Internet, the computer has to understand natural language, determine the answer to a question (or, in the case of Jeopardy, the question to an answer), and then calculate the odds that its answer is correct in order to decide whether it is worth buzzing in.

This is directly applicable to “normal” supercomputing workloads, but it is certainly interesting to see how far we can push the limits of software and silicon.

For more info, read the full article here.

Leave a comment

Happy Birthday Ranger

The folks at the Texas Advanced Computing Center [TACC] are celebrating a birthday today.  Their massive cluster system called Ranger just turned 2!  The 579.4TF machine was the first in the NSF “Path to Petascale” program.  It still remains in the top ten of the Top500 list.

We’re proud of the fact that Ranger has been so widely requested and used for diverse science projects,” said Jay Boisseau, principal investigator of the Ranger project and director of TACC. “It supports hundreds of projects and more than a thousand users — and you don’t attract that many projects and researchers unless you’re running a great, high-impact system. Ranger is in constant demand, often far in excess of what we can provide.”

Ranger consists of 15,744 quad-core AMD Opteron processors, housed within a Sun Constellation cluster.  Beginning Feb 4, 2008, Ranger has enabled 2,863 users across 981 unique research projects to run a total of 1,089,075 jobs and 754,873,713.8 hours of processing time to date.

Ranger enabled the open science and engineering communities to address challenging problems in areas such as astrophysics, climate and weather, and earth mantle convection at unprecedented scales,” Muñoz said. “Ranger truly was a vanguard in NSF’s “Path to Petascale” program and is a testament to what can be done when ‘thinking out of the box.’”

Ranger was a first in many aspects of building and integrating large-scale, commodity HPC.  As a result, the lessons learned in operation has made its way back into many projects.  Infiniband technologies, Lustre, OpenMPI, MVAPICH, and Sun Grid Engine have all reaped the benefit of Ranger’s scale.

These software packages have been significantly improved because of the Ranger project, and are now downloaded by people building clusters in other places. Thus, the Ranger project has had a huge impact on other clusters and the science done on those clusters around the world,” Boisseau said.

Congrats to Ranger and the folks at TACC for advancing the deployment of such super-scale machines.  For more info, read the original article here.

Also posted in Compute, Computing Research, Datacenter operations, HPC Hardware | Leave a comment

Andrew’s Corner: Scientific Codes and Their Creators

In this month’s installment of Andrew Jones’ ZDNet series on high performance computing, he discusses a subject near and dear to my heart: scientific code development.  Given that I’m a classically educated software engineer, I have a somewhat myopic view of code development for scientific computing [et.al, HPC].  Requirements gathering, numerical analysis, code development, rigorous testing and constant evaluation have become a normal part of my life.

Enter Dr. Reg Ular Scientist.  Dr. Scientist takes a different approach.  Does it function and produce reasonable results?  Done.  I realize I’m stereotyping in the worst of ways, but we’ve all seen this happen.  Scientists are, by nature, not software architects.  They’re domain specialists.

By implication, this state of affairs might be responsible for much of what some people see as the mess we are in with respect to assurance of results from the models and portable performance of the codes. The same argument might also be extended to engineering packages and data modelling.

Now that we’ve identified the obvious, how do we strive for better scientific software?  Step one, education.  Educate the scientific masses in the same way that they were originally instructed in their respective discipline.  HPC resources are scientific tools that require careful manipulation and, in many cases, calibration.

Any reputable physical experiment would ensure the instruments are appropriate to the job and have been tested. They would be checked for known error behaviour in the parameter regions of study, and chosen for their ability to give a satisfactory result within a useful timeframe and budget. Those same principles should apply to a software model.

And, of course, collaboration is helpful.

The trick then must be to ensure the scientist code developer understands the methods of numerical software engineering, as well as its issues. Software engineers on the team must equally understand that the code is just part of the science, and not usually a goal in its own right.

As always, Andrew’s article hits the spot.  This is a subject that I face on daily basis.  Ideally, we need to avoid a situation where scientists are forced to become code wizards and conversely, software gurus are forced to become code janitors.  Take a read of Andrew’s latest article here.

Also posted in Computing Research, HPC Software | Leave a comment

DoE Dishes Out Allocations

The US Department of Energy announced today that they have laid out supercomputing allocations totaling over 1.6billion hours to 69 cutting-edge projects.  The allocations come through the DoE Innovative and Novel Computational Impact on Theory and Experiment [INCITE] program.  The INCITE program grants large buckets of HPC resources to groups working on the most challenging of research.  The goal being to drive research results with immediate resources.

Computation and supercomputing are critical to solving some of our greatest scientific challenges,” said Secretary Chu. “This year’s INCITE awards reflect the enormous growth in demand for complex modeling and simulation capabilities, which are essential to improving our economic prosperity and global competitiveness.”

The 69 projects were awarded time at the DoE’s leadership computing facilities [LCF's].  Namely, Argonne National Laboratory and Oak Ridge National Laboratory.  This iteration’s projects include:

  • Energy, including advanced systems for fusion energy and nuclear power, and improving combustion to increase efficiency and reduce emissions to develop safe and renewable energy solutions.
  • Environment, highlighting research into carbon sequestration, developing better insight of natural phenomena like earthquakes and hurricanes, and developing near-zero-emissions combustion devices.
  • Climate change, featuring projects to improve climate models, understand global warming, study the effects of turbulence in oceans, and simulate clouds on a global scale.
  • Biology, including understanding protein membranes to improve drug discovery, diagnostics and better treatment of diseases.

For more info on the various projects and the INCITE program, check out their website here.

Also posted in Computing Research | 1 Comment

ANL researchers developing application to study nuclear reactor cores in action

Got a pointer over the email transom today to news of a new physics application being developed at Argonne National Laboratory. The neutron transport code, called UNIC, is giving researchers their first look at a highly detailed description of a nuclear reactor core.

Elevation plot of the highest energy neutron flux distributionsThe code could prove crucial in the development of nuclear reactors that are safe, affordable and environmentally friendly. To model the complex geometry of a reactor core requires billions of spatial elements, hundreds of angles and thousands of energy groups—all of which lead to problem sizes with quadrillions of possible solutions.

…“The UNIC code is intended to reduce the uncertainties and biases in reactor design calculations by progressively replacing existing multilevel averaging techniques with more direct solution methods based on explicit reactor geometries,” said Andrew Siegel, a computational scientist at Argonne and leader of Argonne’s reactor simulation group.

Although still in development, the code has been run at very large core counts and has already produced new scientific results

In particular, the Argonne team has carried out highly detailed simulations of the Zero Power Reactor experiments on up to 163,840 processor cores of the Blue Gene/P and 222,912 processor cores of the Cray XT5, as well as on 294,912 processor cores of a Blue Gene/P at the Jülich Supercomputing Center in Germany. With UNIC, the researchers have successfully represented the details of the full reactor geometry for the first time and have been able to compare the results directly with the experimental data.

More in the story linked above.

Leave a comment

NVIDIA moves to lock in life sciences

nVidia logoWe’ve had more than ample opportunity to comment on NVIDIA’s GPUs here over the past year or so, and the point I keep coming back to is that they have been genius at building an ecosystem around their product that has created a network effect — the more people that use NVIDIA’s gear, the easier it is for still more people to come into the fold.

Today NVIDIA continues that strategy in the life sciences vertical with the announcement of the Bio Workbench. With the Workbench NVIDIA went out and rounded up 11 popular life sciences applications, made sure they could take advantage of NVIDIA’s GPUs with CUDA, and launched a community site around them. According to that site the codes that are already done include names you’ll probably recognize — AMBER, GROMACS, HOOMD, LAMMPS, NAMD, TeraChem, VMD — with others such as GROMOS, GPU-HMMER, and CUDA-SmithWaterman “coming soon.”

As with GPU-enabled applications in other domains, when they work, they work

“We are working on a new GPU-based technique in the VMD molecular dynamics visualization software that investigates how small molecules like oxygen and CO2 migrate inside proteins.  This research is critical in the study of enzymatic reaction mechanisms,” said John Stone, senior research programmer, University of Illinois at Urbana-Champaign. “A simulation that takes 1 day to run on a GPU-based  workstation would have taken 30 days to run on a CPU-based machine,  rendering it impractical for real research.”

In the life sciences scientists are often trying to use numerical models to create beneficial agents without the mess and expense of experiments in wet labs, generally the point of a lot of numerical modeling no matter what the field. But life sciences results are often a lot more visceral than those of your friendly-neighborhood numerical magnetohydronamicist. For example, researchers at Temple University are using HOOMD from the Bio Workbench codes to improve soaps and shampoos, and talking up the advantages they are seeing over BlueGene/L

Improving the cleaning power of shampoos and liquid detergents and making them more environmentally friendly is as much a computer problem as it is a balance of chemicals. By harnessing the parallel processing power of NVIDIA Tesla GPUs, researchers at Temple University are developing a computer simulation model which provides companies like Procter and Gamble with a fast, cost effective and accurate tool for research and development of surfactant molecules.

…”The computer models needed to accurately simulate surfactant properties are extremely demanding in terms of computational power,” said Axel Kohlmeyer of the Institute for Computational Molecular Science at Temple University. “We discovered that by adding just two NVIDIA Tesla C1060 GPUs, each node in our newest cluster can do 16 times more work, and thus multiplies our local compute capacity far beyond what we could previously get through the national supercomputing centers.”

“To put this into context, we can run a single GPU-optimized molecular dynamics simulation on two Tesla GPUs as fast as we can on 128 CPU cores of a Cray XT3 supercomputer or on 1024 CPUs of an IBM BlueGene/L machine with conventional software,” continues Dr. Kohlmeyer. “With the NVIDIA Tesla GPU-based solution, we now have a more powerful, cost-effective solution that will enable us to advance critical research at a much faster pace. We’re moving rapidly ahead to deploy a larger Tesla GPU cluster at Temple, which will give another huge boost to our work.”


Also posted in Enterprise HPC, GPUs, HPC Hardware, HPC Software | 1 Comment

LANL’s Top 10 Science Stories of 2009

Los Alamos National Lab has released a list of, what they consider, their top ten laboratory science stories of 2009.  The list was based on global viewership of online media content and major programmatic milestones.

Often our top breakthroughs in terms of scientific impact are also the ones that garner the most attention in the media,” said Terry Wallace, Laboratory principal associate director of science, technology, and engineering. “This was certainly the case for Roadrunner and for the Ardi discovery. Sometimes, the best measure of impact is programmatic, such as the successful DARHT two-axis hydrotest, or our teams using nanotechnology for energy breakthroughs. In combination, this collection of advances points to the diverse capabilities at Los Alamos that we harness for national security science.”

So, what made the list!?

  1. Roadrunner: “The Roadrunner supercomputer at Los Alamos is the first computing system in the world to reach a petaflop, computer jargon for 1 million billion calculations per second, a record that stood for a year and a half. But the real accomplishment is that Roadrunner reached that goal using an entirely new computing architecture.”
  2. Ardi: “A Los Alamos National Laboratory geologist is part of an international research team responsible for discovering the oldest nearly intact skeleton of Ardipithecus ramidus, who lived 4.4 million years ago.”
  3. Climate modeling and monitoring: “LANL innovations in high-resolution climate modeling and monitoring led to new insights into the impacts of climate change at global and regional scales.”
  4. MagViz: “LANL’s MagViz team pioneered the use of modified magnetic resonance imagery (MRI) technology to distinguish and alert airport security staff to potentially dangerous liquids and gels in airport carry-on baggage.”
  5. First dual-axis hydrodynamic test: “LANL scientists and engineers fired the first-ever double-viewpoint, multiframe hydrodynamic test at DARHT, the Laboratory’s Dual Axis Radiographic Hydrodynamic Test facility – leading to future experiments at LANL and across the nation’s nuclear security enterprise, supporting the stockpile stewardship and weapons assurance mission.”
  6. Hurricane Prediction: “A system of sensors developed by Los Alamos National Laboratory for the National Nuclear Security Administration’s nonproliferation mission has also begun to give meteorologists their most detailed view of the relationship between hurricanes and lightning.”
  7. Fuel from plants: “Los Alamos National Laboratory has teamed with Solix Biofuels, Inc. to use an award-winning LANL sound-wave technology to optimize production of algae-based fuel in a cost-effective, scalable, and environmentally benign fashion.”
  8. IBEX: “The invisible structures of space are becoming less so, as scientists look out to the far edges of the solar wind bubble that separates our solar system from the interstellar cloud through which it flies.”
  9. Laser-particle acceleration for cancer therapy: “Laser-particle acceleration is an emerging area of physics expected to enable significant future advances in cancer radiotherapy. An international team of physicists led by LANL has accelerated protons to world-record high energies that are otherwise only achievable with large accelerator facilities. Proton radiation at the achieved energy range can be used, for example, to treat eye cancer.”
  10. Nanotechnology for Energy Frontiers: “Two LANL teams were awarded lead roles as DOE Energy Frontier Research Centers to develop new materials for energy.”

For more info on each science topic, read their full article here.

Also posted in Computing Research | Leave a comment

Video Archive

insideHPC.com is a production of insideHPC, LLC. © 2006-2013 Sitemap