作者: T. Valtonen , T. Nurmi , J. Isoaho , H. Tenhunen
DOI: 10.1109/ISCAS.2002.1010495
关键词: Very-large-scale integration 、 Scalability 、 Replication (computing) 、 Distributed computing 、 Embedded system 、 Interconnection 、 Fault tolerance 、 Network topology 、 Synchronization (computer science) 、 Engineering 、 Network on a chip
摘要: In this paper we propose an interconnection scheme for the autonomous error-tolerant (AET) cell introduced in a by Valtonen et al. (2001). The objective here is to partition system into identical, physically and highly configurable cells that can operate without outside control or synchronization. billion transistor Network-on-Chip (NoC) circuits, AET fabric could prove flexible reliable, allow low replication costs, due homogeneousity. However, many challenges persist before useful be constructed: need (i) scalable long-distance communication-global bus wiring fits poorly homogeneous, symmetric limits scalability, (ii) communication-cells all directions, within given range, should accessible (iii) self-synchronizing topologies, absence of external synchronization; (iv) interconnect implemented using near-future technology generations, (v) efficient when constituting billions cells.