Wolfram Research, Inc.

Increasing Performance for Complex Mathematical Modeling

Wolfram Research developed Mathematica, the tool of choice at the frontiers of scientific research, in engineering analysis and modeling, in technical education from high school to graduate school, and wherever quantitative methods are used.

Business Issues

Improve modeling performance
Recommend a scalable platform
Provide 64-bit support with increased memory capacity

Solution

Wolfram Research migrated Mathematica to Solaris 10 and Sun x64 workstations and servers as an ideal 64-bit platform for modeling.

Business Results

Provides a powerful 64-bit modeling system
Increases scalability for grid computing
Offers higher-resolution images
Migrated in half the expected time

Products / Solutions

Success at a Glance

In virtually every industry, complex mathematical modeling activities require leading levels of performance, scalability, and precision. This demand led Wolfram Research to bring its award-winning Mathematica system to Solaris 10 running on Sun x64 systems featuring AMD Opteron processors. "We felt strongly that Mathematica running on the Solaris 10 x64 platform would be a perfect package for anyone looking for speed, dependability, consistent results, and ease of use in development on x64 platforms," explains Joy Costa, Manager of Partnership Programs at Wolfram Research. "We knew it would be a winning combination."

Mathematica 5.2 running on Solaris 10 and Sun x64 platforms provides the perfect blend of performance, scalability, and memory capacity for power-hungry users. For Mathematica, in addition to getting access to the excellent floating-point performance of the AMD Opteron processor, 64-bit support means considerably larger system memory capacity. Without the arbitrary limitations of 32-bit systems, Mathematica models can utilize much larger memory resulting in much higher resolution images with fewer artifacts. With Mathematica’s multi-threading support and Sun’s multi-processor workstations and servers, activities can be performed in parallel by splitting threads or processes among different processors, resulting in considerable scalability. Optional grid Mathematica takes the scalability concept even further, spreading complex tasks across multiple networked systems, each with multiple single- or multi-core processors.

“ I am urging Mathematica users who want a powerful development or production system to move to Solaris 10 on Sun’s workstations and servers based on the AMD processor. ”

— Dr. Steven Christensen, Co-founder, MathTensor, Inc.

Dr. Steven Christensen of MathTensor was engaged to perform testing. "Mathematica running on Solaris 10 is extremely fast and stable," says Christensen. "Solaris has been using 64 bit technology for years now, so there’s no question about how well the OS uses it. It's my view that Mathematica users will find that Solaris 10 and Sun’s x64 systems will prove superior to any other platform."

Wolfram Research was also pleased with the support they received from Sun. "Sun was extremely responsive, quickly supporting every need we had for equipment, engineering assistance, and marketing in a very prompt manner," adds Costa. "The fact that we are already optimized on the newest version of one of the most established OS platforms on the market, speaks well of both Mathematica and Solaris 10. It's really a perfect union."