System Reliability Depends on a Reliable Interconnect

By Tom Cox, RapidIO Trade Association

Too often discussions of COTS deal only with the issue of whether the mechanical system is rugged enough to survive in harsh environments.  Let’s assume the mechanical engineers have made their contribution. COTS using xTCA systems still has several important architectural challenges for the system designer. An obvious area of concern is the backplane and switch fabric interconnect!

If the system cannot perform its basic tasks with sufficient reliability and recovery, then choices of form factor, sheet metal design, and connectors simply do not matter. There are six key requirements for system-level reliability or fault tolerance: No single point of failure, No single point of repair, Fault recovery, 100% fault detection, 100% fault isolation, and Fault containment. Software and system architecture are the key elements in achieving them.

Design engineers have traditionally relied on proprietary methods to meet these requirements. However, as companies rely more on off-the-shelf technology, industry standards must provide the underlying features. Surely the system interconnect must play its role.  The system cannot be reliable or fault-tolerant if the interconnect architecture is not. 
The features required of fault tolerant systems must be embedded into their underlying hardware technologies. The RapidIO interconnect is tailor-made to suit the needs of such demanding applications. And there’s a reason why. The architects of RapidIO’s fault tolerant features (hailing from companies such as Cisco, EMC, IBM, Lucent, Mercury Computer, and Motorola) were experts in high-availability systems and applications.

When targeting your next design for the growing COTS markets in military, defense, aerospace, telecom, industrial, and other applications, you must satisfy their demands for system reliability.  So you should use a switch fabric created with reliability in mind!   That fabric is RapidIO. It was designed specifically for the requirements of industries in which high availability and fault tolerance are major criteria. 

Tom Cox is Executive Director of the RapidIO Trade Association.  You can reach him at tom.cox@rapidio.org. 

RapidIO is a trademark of the RapidIO Trade Association.