A fault-aware low-energy spare core allocation in networks-on-chip

This paper proposes a fault-aware and low-energy technique to spare core allocation in mesh-based networks-on-chip. The proposed technique sets the place of spare cores among free non-failed processing cores, dynamically. Here, dynamically setting means that the places of spare cores are tuned for each application and not fixed in the platform statically. The analysis of a key point leads to introduce the proposed technique: Some tasks of each application can be known as critical, based on their vulnerabilities, the performance degradation, and the energy consumption overheads due to negative impacts of failure recovery. The proposed technique locates the spare cores near to the cores assigned to critical tasks. The results of 1,000,000 fault injection experiments show that the proposed technique leads communication energy reductions and performance improvement, compared to well-known related works.