HPC Applications

HPC Applications

There are a wide range of applications that can be run on a high performance cluster. HPC applications are generally divided into three areas:

• Processor Intensive
• Memory Intensive
• I/O Intensive

The type and use of the application is a major consideration in determining the components and configuration of the cluster.

Processor Intensive
Processor Intensive:
Processor intensive applications require the processing of complex computations. These programs are written to perform high speed mathematics with complex algorithms. Processor intensive applications are used in the video gaming industry, the energy industry and engineering disciplines such as seismic and CAD/CAM applications.

Memory Intensive
Memory Intensive:
Applications requiring access to large datasets such as simulation programs are memory intensive. The system needs adequate memory (RAM) for temporary retention of data sets during processing. Large memory banks reduce the need to access storage devices on a frequent basis during processing. Having a greater amount of RAM available reduces the need for higher levels of throughput and allows faster completion of the processes.

I/O Intensive
I/O Intensive:
I/O (Input/Output) Intensive applications are highly transactional in nature. The application is designed to process a large volume of queries and require multiple, fast connections to the cluster nodes. Finance companies processing credit card transactions, internet portals, online applications or any other applications requiring a high volume of data searches are examples of I/O Intensive systems.

 

Common Types of High Performance systems
Common Types of High Performance systems
Computer clusters can be designed for different uses but generally fall into three categories: High Performance Computing, Load Balancing and Fail-over. (LINKS TO: HPC What is HPC Types of Clusters)