The market for high performance computing (HPC), comprising servers, software, storage, and services, was estimated to be $18.6 billion in 2007 and is expected to grow to $27.2 billion by 2012 according to various industry estimates. Clearly, even in today’s depressed market conditions, companies continue to find significant competitive advantage in applying supercomputers to design, development, and other analytics issues.

Formerly confined to scientific research, HPC is increasingly being applied to solve critical business issues across a wide range of industries ranging from the life sciences to financial markets. HPC technologies also permit resource virtualization of servers, storage, memory, and operating systems as an effective means to increase resource utilization in an enterprise or cloud environment.

Cluster computing has become an attractive alternative to large monolithic commercial computers in many high performance computing applications. It integrates off-the-shelf commodity computers and resources through hardware, networks and software. The main benefits are scalability, availability, affordability, and performance.

However, there is a wide choice available to buyers of cluster computing technology. At the top end, for instance, in the Top50 of the Top500 supercomputer list, there are several systems with exciting architectures based on ultra scale computing (e.g. IBM’s Blue Gene) or special purpose accelerators such as graphics processing units (GPUs).

Vendors are differentiated by their ability to:

• support the transition and growth of multi-core cluster applications
• improve cluster storage performance and power and cooling technology
• improve reliability and availability
• reduce the total cost of ownership in the data center
• provide effective cloud offerings for HPC.
  
  

• Buyers of IT looking to optimize their purchase decisions
• IT vendors looking for competitive analysis, market positioning
• Law firms, venture capital firms, and private equity firms looking to understand the implications of technology

Why Cabot INSIGHTS

Implementing high performance computing in an organization is a complex decision that can cost anywhere from the $120 million IBM RoadRunner to workgroup supercomputer systems that sell for less than $100,000. Understanding HPC requires a level of expertise and insight not easy to find outside large IT vendor companies, that is, the very people who have an interest in selling it.