Available C++ Libraries FAQ

Commercial libraries (A-N)

MtxVec - An object oriented numerical library for Cbuilder

MtxVec is an object oriented numerical library designed for RAD. The library features several new approaches to object oriented numerics and breaks some new ground.

Features:

Encapsulates LAPACK 3.0 and BLAS I,II,III with Intel MKL 3.2.1 Add-on packages available: statistics and dsp. Interfaces Intel NSP for vector operations. Supports SMP for matrix operations. Single and/or double precision supported. Intel Itanium (IA64) ready. Bottom up complete support for complex numbers. Results in 50 to 100aster code then fastest possible structred code (plain procedures) written in C/C++. Design structured in three layers: procedural, object and component; ensures good scalability. Trading space for time technology. Introduces object caching. Trading safety for time technology. Allows user controllable range checking. CPU specific optimizations supporting: PI,PII,PIII. Extreme ease of use and very short time to implement. Ideal for algorithm prototyping. Built in dynamic memory management with extensive support for debugging. Allows completely matrix-vector oriented programming. Requires only CBuilder standard. Support for future CPU instruction sets, without application recompile. (DLL replacement)

Demo (4.5 MB) includes performance and code demonstrations, trial version and complete help system.

A new version 2.0 of MtxVec has been released..

MtxVec adds the following capabilities to C++Builder and .NET development environment:

  • Comprehensive set of mathematical and statistical functions
  • Substantial performance improvements of floating point math by exploiting the P4 SSE2 and SSE3 instruction sets
  • Improved compactness and readability of code.
  • Significantly shorter development times by protecting the developer from a wide range of possible errors.
Typical performance improvements observed by most users are 2-3 times for vector functions, but speed ups up to 10 times are not rare. The matrix multiplication for example is faster up to 20 times. For .NET applications these factors should be multiplied by about 1.3. All comparisons are against classic floating point assembler (!).

Key new features:

  • very fast and clean operator overloading
  • features a very large set of vectorized math functions

Operating Systems

  • PC 32-bit Windows

Compilers

  • Cbuilder
  • Microsoft .NET

Added : 2000-08-15 Amended: 2004-12-17 Licensing : Commercial

Add a comment
Our ads
Buy gold
Beat the credit crunch - buy gold online - quickly, safely and at low prices
www.bullionvault.com
Google
Trumphurst Home Page
C++ Usenet Newsgroup
A Web form for submissions to this list
Contact the author for permission to distribute

Copyright (c) 2024 Nikki Locke, Trumphurst Ltd.
Permission is granted to distribute over the Internet without charge. The author's permission is required (and usually given) to distribute in any other way, including on CD.