Abstract-Today's computers have gigabytes of main memory due to improved DRAM density. As density increases, smaller bit cells become more susceptible to errors. With an increase in error susceptibility, the need for memory resiliency also increases. Self-testing of memory health can proactively check for errors to improve resiliency. This paper describes a software-only selftest to continuously test memory. We present the challenges and design for an approach, called Continuous Online Memory Testing (COMeT), that targets chip multiprocessors. COMeT tests memory health simultaneously with application execution in anticipation of allocation requests. The approach guarantees that memory is tested within a fixed time interval to limit exposure to lurking errors. We developed and evaluated an implementation of COMeT. On the SPEC CPU2006 benchmarks, COMeT has a low 4% average performance overhead. When emulated errors were injected into physical memory, applications executed 1.13x to 4.41x longer with COMeT than without it.

No abstract

No abstract

Memory diagnostics play an important role in addressing the worsening resilience problem for DRAM main memory. As device scales reach the extremes of physical limits, memory is becoming more prone to transient and permanent errors. Online memory diagnostics can be used as part of a comprehensive strategy to mitigate errors. This paper presents a new approach, Asteroid, to incorporate memory diagnostics in a system actively serving a workload. The approach dynamically adjusts itself to workload behavior and resource availability to maximize test thoroughness and minimize performance overhead. In a 16-core enterprise server, Asteroid has modest overhead of 1% to 4% for workloads with low to high memory demand.

No abstract