IBM Journal of Research and Development · 1984 · 13 citations · 9 references
EngineeringComputer EngineeringComputer ScienceFault ToleranceFault-tolerant ComputingFault-tolerant SystemsFault ModelingHardware Fault ToleranceMonte Carlo SimulationReliability ModelingReliable System DesignSimulation MethodsFault SimulationFault-tolerant MemoryMemory ReliabilityMemory System SimulationMonte Carlo Reliability SimulationReliability FunctionSystem ReliabilitySystem Reliability Analysis
Memory systems in modern computers employ a variety of methods to achieve fault tolerance, such as single- or double-error correction, page deallocation, or the use of spare chips or cells. Such methods ensure that the failure rate of the system is considerably less than the sum of the failure rates of the components. However, these methods also complicate the task of evaluating system reliability. The memory reliability function is too intractable to handle analytically, and we must turn to Monte Carlo methods. This paper describes the Fault-Tolerant Memory Simulator (FTMS), an interactive APL program which uses Monte Carlo simulation to evaluate the reliability of fault-tolerant memory systems.
9