 |
|
Maple |
 Enhanced Tools for Solution Developement
Performance Enhancements
Performance enhancements are part of every Maple release. Two important changes in Maple 13 are multithreaded performance improvements and element-wise operations. Other efficiency improvements include:
- Overall performance of medium- to large-scale operations is now faster due to improvements in memory management. On average, the changes result in a speedup of 19% on 32-bit processors and about 10% for 64-bit processors.
- Polynomial operations lie at the heart of many symbolic computations and are often performed, even when the original problem does not involve polynomials. Enhancements to polynomial operations in Maple 13 include:
- New heap-based modular multivariate polynomial multiplication and division algorithms have been added to Maple. These new algorithms are faster than the previous multiplication and division implementations.
- A new algorithm has been added for evaluating the greatest common divisors over algebraic number and function fields, which is particularly effective for sparse polynomials.
- The efficiency of lcoeff , tcoeff, and sort has been improved. In some cases, these operations are twice as fast as in previous versions.
- Enhancements to the LinearAlgebra package include several efficiency improvements, including:
- A new efficient solver for rational solutions of an integer linear system
- A fast modular algorithm for solving linear systems over cyclotomic fields
- A new thin option for the SingularValues command results in a marked speedup for LeastSquares solving when the number of rows is much larger than the number of columns, and uses much less memory
- Copying of a Matrix to a result with a shape (such as creating a Matrix with a triangular shape from an existing full-storage Matrix) is faster, which results in speedups in a variety of matrix operations. For example, a Cholesky decomposition operation can be performed 20 times faster than previous versions. This technique also provides a convenient way to filter the data by masking it with a shape.
|
|
|
|
Connectors
