Visit additional Tabor Communication Publications
October 29, 2009
With great power come great I/O issues
Despite the muscle of today's premier supercomputing systems, scaling and I/O often prevent leading software packages from taking full advantage of the latest hardware's potential power. These ancillary tasks take precious time and resources away from studying the fundamental science that impacts our everyday lives.
To address these issues, a team of researchers from Oak Ridge National Laboratory (ORNL), Georgia Tech, and Rutgers University developed the ADaptable I/O file System (ADIOS). ADIOS is an I/O middleware package that has shown great promise with leading fusion codes, scaling up to 140,000 cores for XGC-1, GTC, and GTS, and in astrophysics with the CHIMERA code. Recently, ADIOS made its mark in the field of combustion.
Researchers at Sandia National Laboratory (Ray Grout, Jackie Chen, and Chun Sang Yoo) and Andrea Gruber of SINTEF, the largest independent research organization in Scandinavia, are using the leading combustion code S3D to perform the first direct numerical simulations of reacting jets in cross flow. These transverse jets are a class of flows used in practical applications in which high mixing rates are desirable -- for example, in fuel injection nozzles in stationary gas turbines for power generation or in aero-gas turbines. The gas turbine industry is faced with numerous R&D challenges in adapting conventional hydrocarbon burner designs to operate safely and cleanly with hydrogen-rich syngas.
Though the S3D team was able to fully scale the code to ORNL's entire Cray XT5 supercomputer, known as Jaguar, the scaling of the I/O proved to be difficult. Enter ADIOS.
Besides its ability to scale, ADIOS's BP file format is resilient to failures in the compute nodes and the file system, an attribute that hinted to researchers that S3D's analysis routines would pair well with the ADIOS ecosystem. After consultations with ORNL's Qing Liu and Scott Klasky (a leading ADIOS developer) and Georgia Tech's Jay Lofstead, Chen's post-doc Ray Grout decided to integrate ADIOS as an alternative I/O mechanism for S3D. ADIOS quickly surpassed the networking limitations imposed by the previous I/O stack.
"I appreciate the clean yet capable interface and that from my perspective, it just works," said Grout. "It would have been fairly difficult for us to get this work finished in time for our targeted paper deadline using our previous I/O solution, but now I'm confident that we'll make it thanks to ADIOS and the help of its team."
S3D's newfound scalability and flexible architecture has led a member of the S3D team, ORNL's Ramanan Sankaran, to explore creating a custom I/O transport method to perform special manipulations to the I/O and the data. With this new transport method, no source code changes would be required.
Thanks to the seamless integration of ADIOS into S3D, the middleware's 1.0 release may have already achieved the team's goal of satisfying 90 percent of ADIOS users "out of the box," said Klasky, adding, "It's really exciting to be able to help the top scientists such as Ray and Jackie obtain results more efficiently. These results make a difference in the world and will hopefully make the world a better place to live."
The ADIOS team thanks all of the scientists who have helped make this work possible, including Hasan Abbasi, Julian Cummings, Divya Dinkar, Ciprian Docan, Stephane Ethier, Garth Gibson, Steven Hodson, Scott Klasky, Zhihong Lin, Qing Liu, Jay Lofstead, Xiaosong Ma, Ron Oldfield, Manish Parashar, Norbert Podhorszki, Milo Polte, Alex Romosan, Ramanan Sankaran, Karsten Schwan, Arie Shoshani, Mladen Vouk, Matthew Wolf, Yong Xiao, Weikuan Yu, and Fang Zhang.
For more information, contact Scott Klasky at firstname.lastname@example.org.
Jun 19, 2013 |
Supercomputer architectures have evolved considerably over the last 20 years, particularly in the number of processors that are linked together. One aspect of HPC architecture that hasn't changed is the MPI programming model.
Jun 18, 2013 |
The world's largest supercomputers, like Tianhe-2, are great at traditional, compute-intensive HPC workloads, such as simulating atomic decay or modeling tornados. But data-intensive applications--such as mining big data sets for connections--is a different sort of workload, and runs best on a different sort of computer.
Jun 18, 2013 |
Researchers are finding innovative uses for Gordon, the 285 teraflop supercomputer housed at the San Diego Supercomputer Center (SDSC) that has a unique Flash-based storage system. Since going online, researchers have put the incredibly fast I/O to use on a wide variety of workloads, ranging from chemistry to political science.
Jun 17, 2013 |
The advent of low-power mobile processors and cloud delivery models is changing the economics of computing. But just as an economy car is good at different things than a full size truck, an HPC workload still has certain computing demands that neither the fastest smartphone nor the most elastic cloud cluster can fulfill.
Jun 14, 2013 |
For all the progress we've made in IT over the last 50 years, there's one area of life that has steadfastly eluded the grasp of computers: understanding human language. Now, researchers at the Texas Advanced Computing Center (TACC) are utilizing a Hadoop cluster on its Longhorn supercomputer to move the state of the art of language processing a little bit further.
05/10/2013 | Cleversafe, Cray, DDN, NetApp, & Panasas | From Wall Street to Hollywood, drug discovery to homeland security, companies and organizations of all sizes and stripes are coming face to face with the challenges – and opportunities – afforded by Big Data. Before anyone can utilize these extraordinary data repositories, however, they must first harness and manage their data stores, and do so utilizing technologies that underscore affordability, security, and scalability.
04/15/2013 | Bull | “50% of HPC users say their largest jobs scale to 120 cores or less.” How about yours? Are your codes ready to take advantage of today’s and tomorrow’s ultra-parallel HPC systems? Download this White Paper by Analysts Intersect360 Research to see what Bull and Intel’s Center for Excellence in Parallel Programming can do for your codes.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?
Join our webinar to learn how IT managers can migrate to a more resilient, flexible and scalable solution that grows with the data center. Mellanox VMS is future-proof, efficient and brings significant CAPEX and OPEX savings. The VMS is available today.