A Methodology for Cognitive NoC Design

Abstract

The number of cores in a multicore chip design has been increasing in the past two decades. The rate of increase will continue for the foreseeable future. With a large number of cores, the on-chip communication has become a very important design consideration. The increasing number of cores will push the communication complexity level to a point where managing such highly complex systems requires much more than what designers can anticipate for. We propose a new design methodology for implementing a cognitive network-on-chip that has the ability to recognize changes in the environment and to learn new ways to adapt to the changes. This learning capability provides a way for the network to manage itself. Individual network nodes work autonomously to achieve global system goals, e.g., low network latency, higher reliability, power efficiency, adaptability, etc. We use fault-tolerant routing as a case study. Simulation results show that the cognitive design has the potential to outperform the conventional design for large applications. With the great inherent flexibility to adopt different algorithms, the cognitive design can be applied to many applications.

Temperature dependence of a sub-wavelength compact graphene plasmon-slot modulator

“…This allows for smallest capacitance beneficial from both energy and AC modulation point of view. Anticipated down scaling of photonic components for interconnect communication links [15,16] can lead to high-functional and reconfigurable network-on-chip solutions [16], to solve the I/O bottleneck in computing [17] and enabling cognitive compute architectures [18,19].…”

A Survey of Software-Defined Networks-on-Chip: Motivations, Challenges and Opportunities

“…Some works described the concept of a cognitive network that can observe, act, learn, and optimize its performance [ 38 , 39 ]. This concept imposes management and supervision requirements on NoC-based systems.…”

Network-Cognitive Traffic Control: A Fluidity-Aware On-Chip Communication

“…In uniform traffic, a node transmits a packet to any other node with equal probability. In hotspot traffic, uniform traffic is applied, but 10% of packets change their destination to one of the following four selected nodes ((7, 2), (7, 3), (7,4), and (7, 5)) with equal probability. In transpose traffic, a node at (i, j) always sends packets to a node at (j, i).…”

“…With the increasing number of processor cores, the communication complexity over the NoC network increases accordingly. A cognitive communication system [3], which is cognizant of the rapid changing in-network traffic requirements promises great performance enhancement for on-chip communication [4], over conventional designs. For example, Melo et al [5] analyzed the router behavior for detecting signal upset in order to minimize error propagations, Chang et al [6] provided a contention prediction scheme for better adaptive routing path decisions, and Tang et al [7] implemented a congestion avoidance method for NoC, based on the speculative reservation of network resource that was proposed by Jiang et al [8].…”