Launched this month, Wolfram Research's Mathematica 7 offers an easy-to-use system for algorithm development, modeling, simulation and visualisation in a single integrated package.
Long overdue in this version is the option to perform parallel high-performance computing. As processor manufacturers move to multi-core architectures, the ability to process complex computations and distribute them over multiple cores is a must.
The software can also be deployed over Mathematica networks deployed across clustered systems and grids.
Wolfram has added 500 new functions and 12 new application areas in version 7, which is supported for Linux x86, Mac OS X, Windows 2000/XP/Vista, Windows Server 2003/2008 and Solaris UltraSPARC/x86. All 64-bit versions of Windows operating systems are supported, as well as other 64-bit operating systems.
We installed the package on XP Professional and openSUSE 11.0 after downloading Windows and Linux versions from Wolfram's servers. Licensing is either standalone or through server licences for specific user groups.
The somewhat idiosyncratic language can be difficult to get hold of initially, but repays deeper investigation, which is necessary anyway to get the best out of the software.
The package works at all levels of expertise, whether you're just browsing through the many application areas, or performing technical research and need to shortcut the time required to perform advanced visualisation and simulation. You can 'reinvent the wheel' if you like, or choose to deploy the many functions to shortcut your specific research area.
Mathematica 7 is a very good package usable in just about any area where technical computing is required, from financial systems to physical and life sciences.