"This version 7.0 release of the IMSL C Library is one of our most ambitious product updates in recent years,â said Dr. Ed Stewart, IMSL Product Manager, Visual Numerics. âThe performance improvements and new analytic capabilities should benefit both our extensive user community and developers looking for new analytic solutions.â
Recent articles and analyst reports document the strong revenue and unit sales growth for multi- and many-core systems. In September, the research firm IDC reported 10% growth for high-performance multi-core systems for the second quarter of 2008 alone. These high-performance systems though pose a challenge for software developers who need to transition from programming for single-core processors to multiple cores. To help alleviate this challenge, Visual Numerics has parallelized numerous algorithms in the IMSL C Library version 7.0 release using OpenMP technology. By using these parallelized algorithms, developers can write applications that take best advantage of the performance capabilities of multi-core systems.
âDevelopers can increase business productivity and keep costs down by taking advantage of the full performance of multi-core technology in high-performance computing applications,â said Ed Turkel, manager, product marketing, Scalable Computing and Infrastructure organization, HP. âThe combination of HP BladeSystem c-Class blade servers and HP ProLiant rack-based servers with the algorithms provided by Visual Numerics allows customers to run their applications with increased performance, while lowering costs and conserving power.â
âWe are pleased to see IMSL C Library multi-core enablement for Windows HPC Server 2008," said Vince Mendillo, director of HPC marketing at Microsoft. "This offering extends Microsoftâs developer ecosystem to leverage the latest high performance computing technology for concurrent computing, while taking advantage of Visual Numerics algorithmic technology.â
The many new and updated algorithms in the IMSL C Library version 7.0 provide unique numerical analysis techniques to customers solving analytic problems in finance, business intelligence, data mining and other areas. For example:
- Finance customers will benefit from a new Feynman-Kac algorithm that solves Black-Scholes problems; a powerful Genetic Algorithm solver to solve problems such as identifying the best technical indicator from hundreds or thousands of indicators that will best predict future stock behavior; and NaÃ¯ve Bayes classification techniques to, for example, classify financial news feeds to text mine for information.
- Business intelligence and data mining software developers can also leverage Genetic Algorithms for optimization and NaÃ¯ve Bayes for classification and text mining problems. In addition, new classification capabilities in the Neural Network algorithm and new ways to select Auto_ARIMA models offer additional classification and forecasting techniques.
- Other new and updated functions, such as faster normal random number generation and new Kochanek-Bartels Cubic Splines, Non-central chi-square, and Non-central studentsâ T PDFs will benefit customers in finance, BI and many other industries.