 |
| | 
| White Paper
Library |
| |
|
| |
Breakthrough Performance for MATLAB®, Python and Other Desktop Applications!
- 10-100X Faster Run Times
- 10-100X Bigger Data Sets
- No Need for Parallel Re-Programming in C/Fortran/MPI
The Star-P® open software platform is delivering revolutionary results to scientists, engineers,
and analysts by enabling them to transparently use high performance computing resources, using
familiar desktop tools. Discover how Star-P users are experiencing a 10X-100X performance increase for their algorithms and applications using Star-P.
|
|
As simulation replaces physical testing, and increasingly complex phenomena are modeled, high performance computers are growing in importance for science and engineering. Interactive desktop science and engineering tools - such as MATLAB®, Python, R, and others - are critical in new model and algorithm development. The current HPC workflow is fundamentally flawed, due to its inability to support an interactive discovery process, but a new programming model is showing promise in enabling desktop tools to operate interactively on servers, clusters, and grids.
Click for access >
_________________________________________________________________________________________________________
|
|
The paper explores some of the performance benefits of Star-P on commodity scalable systems such as IBM's Linux clusters based on multi-core Intel Xeon processors. The results demonstrate substantial performance gains with almost no programmer effort-roughly a 24-fold speed improvement for solving linear matrix equations. An overview of parallel computing with Star-P, a description of the performance test cases and description of IBM cluster configurations used for testing are also addressed.
Click for access >
_________________________________________________________________________________________________________
|
|
As engineers and scientists explore parallel programming approaches, an increasing array of choices is available to them. What is best for your next project? The optimal choice is driven by several factors, including ease of use, scalability, platform scope, and vendors' market focus. To help you sort through these factors, the paper outlines the key questions to ask your potential vendor.
Click for access >
_________________________________________________________________________________________________________
|
 |
The paper outlines the basic steps and tools involved in the process of migrating a desktop application to a parallel environment. It presents typical timelines for going from serial to parallel computing with Star-P based on customer experiences, covering topics such as the learning curve, analysis of the serial application's key opportunities for parallelization, and the practical application of the Star-P tools and methodology.
Click for access >
_________________________________________________________________________________________________________
|
 |
This case study examines the usage of Star-P at Air Force Research Lab, one of the early adopters of this product as well as a customer that has already benefited from the implementation of the product. IDC explores the conditions of the facility, motivations of the product adoption, the results of the implementation, and the benefits and challenges of the solution.
Click for access >
_________________________________________________________________________________________________________
|
 |
Star-P enables quantitative analysts using desktop tools such as MATLAB® to operate interactively with parallel computer systems, eliminating the need for re-programming, and accelerating time-to-solution. Analysts preserve their familiar financial engineering workflow, while gaining a quantum leap in performance - for more accurate forecasting, improved risk analysis, more effective trading strategies, and other requirements involving complex algorithms and massive databases.
Click for access >
_________________________________________________________________________________________________________
|
 |
The application of Star-P to semiconductor design and manufacturing is illustrated through several examples:
- Accelerating Monte Carlo SPICE simulations
- Static timing validation
- Running large simulations from desktop SPICE
- Supporting "Design for Manufacturability" (DFM) methodologies
Click for access >
_________________________________________________________________________________________________________
|
 |
The report summarizes a recent study of the current state of custom application development in high performance technical computing. Based on hundreds of surveys and interviews, the report examines the software tools currently used among several industries, probes current application development environments, practices, and limitations, and catalogs critical issues and bottlenecks.
Click for access >
_________________________________________________________________________________________________________
|
 |
The report summarizes a recent study of the current state of technical computing with the open source Python language. Based on hundreds of surveys and interviews, the report examines the application areas, types of computation, user characteristics, performance issues, and unmet needs.
Click for access >
_________________________________________________________________________________________________________
|
 |
[Originally published in HPCwire on 12/5/2005]
This article focuses on the emerging trends in the development of parallel applications for high performance technical computers (servers, clusters, and grids). It summarizes the state of the industry today, explores how the industry may evolve over the next several years, and presents some choices available for engineers, scientists, and analysts trying to maximize the return on their parallel computing investments.
Click for access >
_________________________________________________________________________________________________________
|
 |
Ron Choy and Alan Edelman, Proceedings of the IEEE, Volume 93, Issue: 2, February 2005.
MATLAB® is one of the most widely used mathematical computing environments in technical computing. It is an interactive environment that provides high performance computational routines and an easy-to-use, C-like scripting language. In a recent survey, 27 parallel MATLAB® projects have been identified. This paper expands upon that survey and discuss the approaches the projects have taken to parallelize MATLAB®, describes the innovative features in some of the parallel MATLAB® projects, and concludes with an idea of a ‘right’ parallel MATLAB®.
Click for access >
_________________________________________________________________________________________________________
|
 |
By Bill Blake, Interactive Supercomputing, Reprint from Scientific Computing, March 2007
Shifting the Focus to Science: New programming model allows researchers to utilize multi-core systems without manual re-writes.
Click for access >
_________________________________________________________________________________________________________
|
|
 |
| |
|
|
 |
Star-P Support
Blog
Email
Page
