Dynamic reliability management adapts voltage limits to a processor’s workload and accumulated wear. By modeling key failure mechanisms and adjusting dynamic voltage scaling, system designers can use reliability margin to improve performance during periods of high demand.