Networks-on-chip need to survive to manufacturing faults in order to sustain yield. An effective testing and configuration strategy however implies two opposite requirements. One one hand, a fast and scalable built-in self-testing and self-diagnosis procedure has to be carried out concurrently at NoC switches. On the other hand, programming the NoC routing mechanism to go around faulty links and switches can be optimally performed by a centralized controller with global network visibility. To the best of our knowledge, this article proposes for the first time a global network testing and configuration strategy that meets the opposite requirements by means of a fault-tolerant dual network architecture and a fast configuration algorithm for the most common failure patterns. Experimental results report an area overhead as low as 12.5% with respect to the baseline switch architecture while achieving a high degree of fault tolerance. In fact, even when multiple stuck-at faults are considered, the capability of fault masking by the dual network is always over 80%, and the support for multiple link failures is more than 90% in presence of two unusable links in the main network with minimum set-up times.

Abstract-Current and future on-chip networks will feature an enhanced degree of reconfigurability. Power management and virtualization strategies as well as the need to survive to the progressive onset of wear-out faults are root causes for that. In all these cases, a non-intrusive and efficient reconfiguration method is needed to allow the network to function uninterruptedly over the course of the reconfiguration process while remaining deadlock-free. This paper is inspired by the overlapped static reconfiguration (OSR) protocol developed for off-chip networks. However, in its native form its implementation in NoCs is outof-reach. Therefore, we provide a careful engineering of the NoC switch architecture and of the system-level infrastructure to support a cost-effective, complete and transparent reconfiguration process. Performance during the reconfiguration process is not affected and implementation costs (critical path and area overhead) are proved to be fully affordable for a constrained system. Less than 250 cycles are needed for the reconfiguration process of an 8x8 2D mesh with marginal impact on system performance.

No abstract

No abstract

No abstract