Built-in Parallel Computing: New in Mathematica 7

Built-in Parallel Computing

Mathematica 7 adds the capability for instant parallel computing. On any multicore computer system, Mathematica 7 is automatically set up to be able to run multiple parts of a computation concurrently—and for the first time makes parallel computing easy enough that it can be used in seconds as a routine part of everyday work. The symbolic character of the Mathematica language allows unprecedentedly straightforward support of many existing and new parallel programming paradigms and data-sharing models—and Mathematica's parallel infrastructure is set up to allow seamless scaling to networks, clusters, grids and clouds.

Multicore parallelism standard with zero configuration on all versions of Mathematica. »
Parallelize function for automatic parallelization of Mathematica code. »
Automatic parallel versions of iteration and functional operations. »
ParallelTry function for speculative parallelization. »
Flexible data parallelism functions built directly into the Mathematica language.
Broadcast functions for separate evaluation on multiple subkernels. »
Automatic and tunable division of computations for parallel load balancing. »
Built-in concurrency primitives for full control of concurrent computations. »
Shared memory capabilities for symbols and functions. »
High-level synchronization with critical section support. »
Automatic and manual subkernel management. »
High-level mechanisms for distributing definitions to subkernels. »
In-notebook display of concurrent process status. »
Full user interface for manual configuration of parallel architecture. »
Parallel debugging and profiling through Wolfram Workbench.
Built-in support for parallelism on homogeneous and heterogeneous networks and clusters. (Requires gridMathematica.)
Built-in support for the Wolfram Lightweight Grid System for ad hoc parallel networks.
Built-in support for Microsoft, Sun, LSF, PBS, etc. cluster management systems.

Automatically Parallelize Computations Create Tables of Results in Parallel Solve Optimization Problems with Speculative Parallelism

Easily Distribute Function Definitions in Parallel Monitor the Parallel Efficiency of a Computation Dynamically Display Load Balancing in a Parallel Computation

Automatically Monitor the Progress of Concurrent Computations Related Functions Parallelize . ParallelTry . ParallelMap . ParallelTable . ParallelSum . ParallelProduct . ParallelDo . ParallelArray . ParallelCombine . ParallelEvaluate . DistributeDefinitions . ParallelNeeds . ParallelSubmit . WaitAll . WaitNext . SetSharedVariable . $SharedVariables . SetSharedFunction . $SharedFunctions . UnsetShared . CriticalSection . LaunchKernels . Kernels . $KernelCount . $KernelID . $ProcessorCount . AbortKernels . CloseKernels More About Parallel Computing Parallel Computing Tools User Guide See Also New In 7.0 Integrated Programmatic Email New in 6: Instant High-Level Debugging