ECE498 Dependable Computing Systems

– Sources and Type of failures
– Evaluation metrics and evaluation methods
– Error modeling approaches
– Techniques for error detection
– Fault tolerant mechanisms in processors
– Fault tolerant mechanisms in memories
– Error mitigation mechanisms at circuit and architecture layer
– Mechanisms for addressing failures at the software
– Cross layer error resilient approaches

The course aims at the development of knowledge and skills for the design of dependable computing systems such that they operate efficiently under various operating conditions. After the successful completion of the course, participants will be able to:

• Understand the challenges related to the design of computing systems that operate under conditions that increase the failure probability and endanger the reliable operation.
• Explain the various sources of potential failures and their impact on processors, memories and the software.
• Apply techniques at various layers (with an emphasis at circuit and micro-architecture) for addressing potential failures.
• Understand and evaluate the impact of failures and the error mitigation mechanisms on the performance, power and/or area of the computing system.
• Implement and evaluate error mitigation mechanisms.