ASPEED Software Clears the Path to Optimizing Application Performance

White Paper Maps the 6 Most Common ‘Gotchas’ when Re-Programming for Multi-Processing – and How to Avoid Them

NEW YORK (June 13, 2005) — ASPEED Software is offering a fast-track for organizations needing to "upgrade" crucial business applications to support today’s complex configuration and service needs. Its newest white paper maps the most common pitfalls in multi-processing – and concrete strategies to overcome them.

ASPEED is increasingly considered the key player in the rapidly expanding market for enabling even the most complex computationally intense applications to tap the power of parallel distributed computing.

"Designing Distributed Parallel Programs for Performance" is currently available at It details an application-centric approach that achieves critical IT and business goals with significantly less investment, higher predictability levels and optimal response times.

It analyzes the four most frequently used application distribution techniques to reduce run times: splitting data, re-engineering and decomposing, re-engineering with middleware and applying higher level adapters to the existing application.

"The central issue with application performance is really a measure of three key factors: run time, scalability and performance predictability," said Kurt Ziegler, evp of development for ASPEED. "Making the ‘right’ application changes, whether it’s splitting the application vertically or in parallel, can reduce run times and increase scalability – but it does not ensure predictability. More predictable and scalable run times demand additional considerations that can only be achieved at the application level, and this white paper provides companies with the tools to find the correct alternative when distributing on a grid."

Market Pressure Grows

Across industries, companies are demanding more responsive and timely delivery of near-perfect analysis and results. This pervasive pressure is driving the convergence of three critical priorities:

- optimizing utilization

- leveraging multiple commodity computers and

- delivering faster and more predictable and scalable service levels

As a result, CIOs, architects and IT decision-makers are challenged to re-assess existing application architectures and programming designs to leverage the more flexible and cost-effective computing infrastructures – currently characterized as clusters and grids.

Data from several of the world’s largest financial institutions affirms that the best performance, run-time reductions, predictability and scalability are clearly linked to "algorithm integrity." By taking an approach that retains the original applications’ inherent control and preserves the original flow and algorithm integrity, companies are improving overall application performance and reducing peak capacity demand, while ensuring the most rapid and robust upgrade path.

Unfortunately, many enterprises unwittingly end up spending far more time and money upgrading, while not achieving the application performance, predictability or scalability needed, without gaining a clear path for making the application truly "future proof."

These findings are directly leveraged by ASPEED’s ACCELLERANT software, which has helped some of the world’s top banks and investment firms enable core applications to quickly leverage the benefits of cluster and grid computing, and to quickly achieve predictable faster response times, scalability and robustness for their most critical business applications.

ASPEED offers the IT industry’s only application upgrade approach to enable any single-thread or multi-thread application to enjoy the full benefits of multi-CPU hardware, cluster or grid configurations without major application surgery or re-engineering.

Become a bobsguide member to access the following

1. Unrestricted access to bobsguide
2. Send a proposal request
3. Insights delivered daily to your inbox
4. Career development