Innovative practices session 11C: Resilience
Chen-Yong Cher, Mohan Kumar · 2013
Software Hardware co-design is a key mechanism to realizing reliability at platform level. While error avoidance is the ideal, it is not always practical in an economic sense. Often used technique is to detect errors in hardware and implement the recovery in coordination with software. This talk will discuss some examples of such coordination on Intel Xeon (e.g., DIMM Sparing, MCA recovery) and the implications of this type of solution such as platform validation, interface standardization, software enablement, etc.