October 24, 2013
Oct. 24 -- The Center For Research Computing (CRC) at the University of Notre Dame, South Bend, Indiana has deployed Allinea Software's tools for HPC development as part of their mission to open their HPC resources to a wider user base and simultaneously improve the success and efficiency of codes running on their systems.
Tim Stitt, Head of User Services at the CRC, said, “A critical part of my team’s efforts is invested in supporting our ever-growing user-community – over 1500 user accounts. As more users join, my team needs to help port more codes to CRC resources and ensure resources are used effectively.”
“We rely on Allinea Software’s debugging and profiling tools, Allinea DDT and Allinea MAP, to provide a shallow learning curve and a consistent, easy to use user-interface. The tools give users the power and flexibility to understand intimate details about their code’s performance.”
“By putting the demands of debugging and profiling directly in the students’ hands, my team can focus on those who really need our expert help."
Making fast, efficient research accessible to all
As well as traditional HPC disciplines, CRC supports researchers from the University’s social science and arts faculties.
“Both Allinea DDT and Allinea MAP greatly help students from all faculties. Even with limited computing backgrounds, Allinea Software’s intuitive interface allows them to quickly achieve the results they want,” he says.
Best-in-class tools deliver best in-class experience
Stitt teaches parallel computing programs. His students are exposed to parallel debugging and profiling best-practices that include extensive use of both Allinea DDT and Allinea MAP – in fact, they are essential to the final project for the course.
Stitt comments, “It’s important to teach good debugging and profiling techniques within the curriculum to nurture the next generation of HPC programmers’ natural programming practices. Allinea Software tools are so easy for students to learn.”
“Introducing high quality and easy-to-use tools quickly demystifies debugging and profiling. With Allinea Software’s tools it’s no longer a dark art but an enjoyable experience that gives great results.”
David Lecomber, CEO of Allinea Software, says, “It’s great to see our products helping students take advantage of HPC and learning the benefits of debugging and profiling tools hands-on. These students are doing research they’d never have dreamed of, thanks to Tim and his team at Notre Dame, and we’re proud to be part of that.”
Storm-Surge Simulations more efficient with Allinea MAP
Stitt is collaborating with a high-profile research group on campus that develops a computational model for hurricanes and storm-surges.
The code is used by the US Federal Emergency Management Agency (FEMA) and Department of Defense. It helped design the rebuild of the levees in New Orleans after Hurricane Katrina and develop hindcasts of Hurricane Sandy to improve future forecasting.
Using Allinea MAP, Stitt has been able to convert existing pure MPI code to a hybrid of MPI and OpenMP for the Intel Xeon Phi Stampede machine at Texas Advanced Computing Center (TACC). Allinea MAP helped him quickly identify critical performance hotspots in the existing code that now drives the development of the hybrid code.
“Profiling the code with Allinea MAP gave insights on how to develop a hybrid programming approach and tune it effectively for the Intel Xeon Phi based system. It allowed me to identify two crucial routines that are called frequently within our storm-surge runs and with this knowledge I’m now working to distribute this across multiple cores using OpenMP with the goal of offloading it to the Intel Xeon Phi coprocessors.”
“Initial recoding efforts are already showing significant performance gains and it’s all down to the quick and simple application of the Allinea MAP tool,” he says.
Source: Allinea Software
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
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?