Dynamic data flow analysis for NoC based application synthesis

Abstract

Network-on-Chip (NoC) is an interesting communication fabric for multi processing element architectures that benefits from the parallelism of algorithms. We present a method that uses a symbolic execution technique to extract the parallelism of an application to be mapped on FPGAs using the flexibility of a NoC communication infrastructure and the properties of a high level programming language. An application specific hardware is then generated using a High Level Synthesis flow. We provide a dedicated mechanism for data paths reconfiguration that allows different applications to run on the same set of processing elements. Thus, the output design is programmable and has a processor-less distributed control. This approach of using NoCs enables us to automatically design generic architectures that can be used on FPGA servers for High Performance Reconfigurable Computing. We validate our method on binomial tree applications used for option pricing on FPGAs.

A Software Framework for Rapid Application-Specific Hybrid Photonic Network-on-Chip Synthesis

“…Kapadia et al [34] also proposed a 3-D NoC-PDN cosynthesis framework to minimize communication power in a regular NoC while also reducing worst-case IR-drops in the power delivery network (PDN). Other efforts focus on application specific synthesis of irregular NoCs [35][36][37][38][39][40]. For example, Murali et al [35] present a synthesis technique that utilizes min-cut partitioning to allocate switches to groups of custom cores and minimize NoC power consumption.…”

“…Srinivasan et al [36] present a genetic algorithm based approach to synthesize a low power custom NoC topology. Payet et al [39] present a dynamic approach to data dependency profiling and its associated synthesis chain to generate NoCs. Romanov et al [40] propose using irregular topologies for the synthesis of NoCs based on task graph analysis.…”

“…Within the PSO framework, an individual entity is called a particle and it shares information with other particles, either in the form of direct or indirect communication to coordinate their problem-solving activities. In recent years, the PSO algorithm has been applied to many combinatorial optimization problems such as optimal placement of wavelength converters in WDM networks [38] and reconfiguration of field-programmable analog circuits [39].…”

“…This process continues for the application, until a dominating (global) solution emerges. been applied to many combinatorial optimization problems such as optimal placement of wavelength converters in WDM networks [38] and reconfiguration of field-programmable analog circuits [39].…”