System News
"Sun Business Ready HPC for Altair RADIOSS"
An Assessment of the Cluster Approach to HPC
December 23, 2009,
Volume 142, Issue 4

an overview of running RADIOSS in a high-performance clustered environment on Sun systems
 

In his Sun BluePrints paper, "Sun Business Ready HPC for Altair RADIOSS: Best Practices for Altair RADIOSS Deployments, Dr. Gunter Roeth, of Sun ISV Engineering provides insight into constructing an optimal configuration for Altair RADIOSS car crash simulations using a high-performance cluster of Sun systems equipped with x86/x64 processors. The paper points out how HPC solutions often consist of servers equipped with Intel Xeon or AMD Opteron processors that are clustered together using high-speed, low-latency interconnects. In many cases, Roeth writes, these high-performance clusters can match and even surpass the throughput capabilities of vertically-scalable proprietary solutions.

A global software and services company, Altair focuses on data analysis, visualization, and high-performance computing in business and engineering. Central to these services is the Altair HyperWorks suite of enterprise analytics applications that includes statistical, database, visualization, and simulation software. An open architecture solution, HyperWorks also offers broad interoperability with commercial computer-aided design (CAD) and computer-aided engineering (CAE) solutions for greater flexibility and efficiency.

RADIOSS, a next-generation implicit and explicit finite-element solver for linear statics and dynamics as well as complex nonlinear transient dynamics and mechanical systems simulation, is an important component of the HyperWorks suite. The very size and complexity of RADIOSS models frequently requires the use of the massively parallel computation capabilities that can be delivered by large symmetric multiprocessing (SMP) systems or high-performance compute clusters, Roeth notes.

The author points out that high-end Sun systems with SPARC processors have long been supported for RADIOSS and offer users up to 256 cores of processing power to help speed computations. Alternatively, organizations can create a high-performance compute cluster and take advantage of affordable, energy efficient Sun servers with Intel and AMD processors. In addition to support for the Linux operating environment, a recent software porting initiative lets organizations utilize the Solaris Operating System (OS) for RADIOSS HPC clusters with x86/x64 processors.

Roeth reports on the numerous benchmarks Sun conducted with Altair’s Hyperworks RADIOSS crash simulation using a variety of x86/x64 processors, cluster interconnects, operating systems, and file systems in an effort to determine the best hardware configuration for a RADIOSS implementation. Roeth describes the results of these benchmark tests that were designed to assist organizations with the process of selecting an optimal hardware configuration for RADIOSS workloads.

Roeth's article provides an overview of running RADIOSS in a high-performance clustered environment on Sun systems. Sample performance data and suggestions for designing and configuring optimal solutions for a variety of workload sizes are presented. More specifically, this document addresses the following topics:

  • “Constructing a Compute Cluster for Altair RADIOSS” provides a description of RADIOSS and offers general configuration information.

  • “Measuring and Optimizing RADIOSS Crash Performance” provides benchmark results and guidelines that can help organizations optimize RADIOSS deployments.

  • “Recommended Configuration” includes information on a Sun Business Ready HPC Solution that satisfies the requirements of RADIOSS workloads.

Roeth writes that his article assumes a basic understanding of RADIOSS and the capabilities it provides and that it was written primarily for IT administrators who deploy and configure RADIOSS, as well as for experienced users who are interested in performance-related issues.

The paper's summary of recommended configurations includes the computational needs of both RADIOSS Starter and RADIOSS Engine.

The paper provides a table listing specific technology and product recommendations for each major component of a RADIOSS solution and the associated benefits.

More Information

Sun High Performance Computing

Sun Modular Systems

Sun Fire Servers

Solving the HPC I/O Bottleneck: Sun Lustre Storage System

Sun Business Ready HPC for MD NASTRAN [...read more...]

Keywords:

fullsource
 

Other articles in the HPC section of Volume 142, Issue 4:

See all archived articles in the HPC section.



News and Solutions for Users of Solaris, Java and Oracle's Sun hardware products
Just the news you need, none of what you don't – 42,000+ Members – 24,000+ Articles Published since 1998