Partner Newsletter Q4 2010 - Intel Cluster Ready Articles 3-4

Partner Highlights

Landmark and Appro:
Seismic Coverage Validation Engine for Oil and Gas Exploration Data Processing

Powerful oil and gas exploration tools, simplified deployment. That’s what Landmark, Appro and Intel are delivering as Landmark brings its new Seismic Coverage Validation Tools to market on a certified Intel Cluster Ready cluster from Appro. Customers gain the power of the Landmark validation tools, plus the simplicity of a pre-tested, ready-to-run Appro GreenBlade* System powered by the Intel® Xeon® processor 5600 series.

Landmark’s wave equation-based Illumination and Visibility software helps geoscientists design 3D seismic surveys for geologic targets. It’s well suited to meeting the challenges of complex subsurface environments, like deepwater and subsalt, where data variations can lead to incorrect interpretations and reservoir models. By enabling geoscientists to analyze and model a seismic signal from source, to subsurface target, to receiver, Landmark’s Illumination and Visibility technology allows geophysicists to optimize imaging parameters to a specific subsurface target or make required design changes to future seismic acquisition surveys.

These accurate wave equation-based techniques are compute intensive, so they benefit from the high performance of the Intel Xeon processors and the Intel Cluster Ready certified cluster from Appro. This combination, optimized to run Landmark’s Seismic Coverage Validation tools, provides faster results so geoscientists can make improved and timelier exploration decisions while lowering overall exploration costs.

Read the Solution Brief

See what the Oil and Gas industry is saying about the solution:
E&P: Seismic validation comes prepackaged
Offshore Magazine: Package offers seismic coverage validation

ANSYS Virtual Seminar: Boost your productivity with HPC

Intel Cluster Ready partners, ANSYS and Appro, team up in a webinar that shows how to simulate, analyze, and visualize models faster with HPC and how Intel Cluster Ready helps to ease the traditional pain points of high performance clusters.

Read Brock Taylor’s blog for more
View the recorded webinar. "Available soon."

Intel Cluster Ready News

New Release: Intel® Cluster Checker 1.6
A great tool to simplify cluster verification gets even better

Intel® Cluster Checker, the go-to tool for verifying your clusters’ configuration and performance, is now more powerful than ever with the release of Intel Cluster Checker 1.6 for Linux*.

Intel Cluster Checker 1.6 adds a variety of new features in version 1.6 aimed at regular (not privileged) users. For example, regular users will now be able to check hardware uniformity and I/O performance. Other capabilities were added to improve the user experience. You’ll now find:

• A centralized approach for network fabric selection
• A standard output directory
• An output log analyzer
• The source for an XML parser for output logs
• An experimental mode that allows for reverse execution

Check Multiple Spec Versions in a Single Execution
As the Intel Cluster Ready Specification continues to evolve, Intel Cluster Checker 1.6 is designed to allow you to verify compliance against multiple specification versions in a single execution. In addition, third-party benchmarks have been updated to apply advanced optimizations and leverage new platform support offered by the latest release of Intel® Cluster Toolkit, specifically:

Intel® MPI Library 4.0.1
Intel® Fortran/C++ Compiler 12.0
Intel® Math Kernel Library 10.3

Now Even Faster
Intel Cluster Checker 1.6 delivers significant speedups for different execution modes. A major performance increase came from enhancing the most relevant test modules, including imb_pingpong_intel_mpi test module.

Learn More
Read the 1.6 Release Notes

Read Andres More’s Intel Cluster Checker 1.6 Tips & Tricks

Download the latest release at the Intel® Registration Center and click on Intel® Cluster Checker.

See What’s New with Intel Cluster Tools

By Gergana Slavova

While Intel® Cluster Ready clusters are designed to be ready-to-run black boxes for a variety of ISV applications: just what end-users need to help them take advantage of high-performance computing. They come with the full suite of Intel tools runtimes, free of charge for end users.

Intel Cluster Ready clusters also provide a great base system standard for further cluster development. With the new Intel® Cluster Tools, it’s easier than ever to develop customized clusters and take the next step in performance and scalability. The new versions of Intel Cluster Toolkits, to be released soon, offer updates to the latest Intel® Tools, bundled in a single easy-to-install package, with one license and one support service. Included are the latest versions of:

• Intel® C++ and Fortran Compiler XE
• Intel® MPI Library
• Intel® Trace Analyzer and Collector
• Intel® Math Kernel Library
• Intel® MPI Benchmarks

The brand new Intel® MPI Library is a high-performing implementation of the latest MPI-2.1 standard, enabling applications to run across multiple fabrics via runtime selection. This version provides new levels of performance and flexibility through improved interconnect support, faster on-node messaging, and an application tuning capability that adjusts to the cluster architecture and application structure. Startup times are better than ever with a new process manager, interfacing seamlessly with top industry batch systems. On the collective algorithm level, further optimizations allow for faster runtimes.

Great Support for Developing Hybrid Apps
What about hybrid applications? Whether you’re using Intel MPI with OpenMP, simple threads, or one of the highly parallelized routines in the Intel® Math Kernel Library, you’re set: Cluster Tools supports them all. While the Intel MPI Library selects the best pinning defaults for your processor and cluster structure, it also offers a large number of options, all tunable at runtime.

Come See Our ParaView Demo at SC10
Currently, the Intel Tools team is using the Intel Cluster Tools bundles to explore how ParaView – a scalable, open-source framework for scientific visualization and data analysis – with the right combination of Intel hardware and software, can be more useful to the HPC community than GPU-based solutions. We’ll be showing off the results of this investigation at SC10. Stop by the Intel booth for a hands-on exhibit.

Find out more about the Intel Cluster Tools Suites
Gergana Slavova is a technical consulting engineer at Intel. She earned her BS in Computer Science from the University of Illinois at Urbana-Champagne, and will soon celebrate her fifth anniversary at Intel.


For more complete information about compiler optimizations, see our Optimization Notice.