Cross-Layer Reliability (RelXLayer)

The Cross-layer Reliability (RelXLayer) visioning process addresses the fact that we will no longer be able to reliably design or manufacture fault-free hardware systems. As the critical dimensions of devices, such as transistors and wires, used to implement computer systems shrink to only a few nanometers, rates of transient faults, permanent defects, and variation between devices on the same die are expected to increase to the point where today's fault-tolerant approaches will no longer be practical. Instead, computer systems will need to adopt a model in which each layer in the abstraction hierarchy - applications, O/S, architecture, circuits - is prepared for the layer below to transmit bad data and in which all of the layers in the hierarchy cooperate to deliver correct operation in spite of faults, variations, and other effects. Exacerbating this challenge is the need to continually reduce net energy per operation while providing this protection.

RelXLayer draws participants from the broader computer research community to nurture a vision for a multi-level approach to reliability, generating a clear picture of the challenges and opportunities offered in multi-level reliability approaches. To foster this community, the group held three workshops to engage researchers and practitioners aimed at addressing this abstraction hierarchy and initiating cross-layer discussions.


First Meeting
March 26-27, 2009 at Intel Santa Clara

Second Meeting
July 8-9, 2009 at Los Alamos National Labs

Third Meeting
October 29-30, 2009 at IBM in Austin, TX



Special Session at DATE 2010

2009 IEEE Workshop on Silicon Errors in Logic - System Effects


Andre DeHon (University of Pennsylvania)
Nick Carter (Intel Corporation)
Heather Quinn (Los Alamos Laboratory)