Circuit Switched Run-Time Adaptive Network-on-Chip for Image Processing Applications

Author(s):  
Lars Braun ◽  
Michael Hubner ◽  
Jurgen Becker ◽  
Thomas Perschke ◽  
Volker Schatz ◽  
...  
Author(s):  
Liang Guang ◽  
Ethiopia Nigussie ◽  
Juha Plosila ◽  
Hannu Tenhunen

Self-aware and adaptive Network-on-Chip (NoC) with dual monitoring networks is presented. Proper monitoring interface is an essential prerequisite to adaptive system reconfiguration in parallel on-chip computing. This work proposes a DMC (dual monitoring communication) architecture to support self-awareness on the NoC platform. One type of monitoring communication is integrated with data channel, in order to trace the run-time profile of data communication in high-speed on-chip networking. The other type is separate from the data communication, and is needed to report the run-time profile to the supervising monitor. Direct latency monitoring on mesochronous NoC is presented as a case study and is directly traced in the integrated communication with a novel latency monitoring table in each router. The latency information is reported by the separate monitoring communication to the supervising monitor, which reconfigures the system to adjust the latency, for instance by dynamic voltage and frequency scaling. With quantitative evaluation using synthetic traces and real applications, the effectiveness and efficiency of direct latency monitoring with DMC architecture is demonstrated. The area overhead of DMC architecture is estimated to be small in 65nm CMOS technology.


VLSI Design ◽  
2012 ◽  
Vol 2012 ◽  
pp. 1-17 ◽  
Author(s):  
Emanuele Cannella ◽  
Onur Derin ◽  
Paolo Meloni ◽  
Giuseppe Tuveri ◽  
Todor Stefanov

System adaptivity is becoming an important feature of modern embedded multiprocessor systems. To achieve the goal of system adaptivity when executing Polyhedral Process Networks (PPNs) on a generic tiled Network-on-Chip (NoC) MPSoC platform, we propose an approach to enable the run-time migration of processes among the available platform resources. In our approach, process migration is allowed by a middleware layer which comprises two main components. The first component concerns the inter-tile data communication between processes. We develop and evaluate a number of different communication approaches which implement the semantics of the PPN model of computation on a generic NoC platform. The presented communication approaches do not depend on the mapping of processes and have been implemented on a Network-on-Chip multiprocessor platform prototyped on an FPGA. Their comparison in terms of the introduced overhead is presented in two case studies with different communication characteristics. The second middleware component allows the actual run-time migration of PPN processes. To this end, we propose and evaluate a process migration mechanism which leverages the PPN model of computation to guarantee a predictable and efficient migration procedure. The efficiency and applicability of the proposed migration mechanism is shown in a real-life case study.


Sign in / Sign up

Export Citation Format

Share Document