DocumentCode
3061948
Title
COMeT: Continuous Online Memory Test
Author
Rahman, Musfiq ; Childers, Bruce R. ; Cho, Sangyeun
Author_Institution
Comput. Sci. Dept., Univ. of Pittsburgh, Pittsburgh, PA, USA
fYear
2011
fDate
12-14 Dec. 2011
Firstpage
109
Lastpage
118
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 self test 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 1.13× to 4.41× longer with COMeT than without it.
Keywords
DRAM chips; multiprocessing systems; program testing; COMeT; DRAM density; SPEC CPU2006 benchmarks; allocation requests; chip multiprocessor; continuous online memory testing; error susceptibility; lurking error; memory health self-testing; memory resiliency; performance overhead; software only self-testing; Instruction sets; Kernel; Linux; Memory management; Random access memory; Resource management; Testing; Self testing; errors; memory; resilience;
fLanguage
English
Publisher
ieee
Conference_Titel
Dependable Computing (PRDC), 2011 IEEE 17th Pacific Rim International Symposium on
Conference_Location
Pasadena, CA
Print_ISBN
978-1-4577-2005-5
Electronic_ISBN
978-0-7695-4590-5
Type
conf
DOI
10.1109/PRDC.2011.22
Filename
6133072
Link To Document