2010Unpublished venueRequires access

NoC datapath for polymorphic processors in embedded systems

Joshua Weber, Erdal Oruklu, Jafar Saniie

Open publisher page 1 citations

Abstract

Polymorphic processing has the goal of producing a processor combining the advantages of general purpose processing with the significant gains achievable by custom application specific computing. To achieve these ends a novel polymorphic processor architecture is presented. Incorporating networking on a chip (NoC) techniques into the datapath design has the potential to provide noticeable advantages when compared to a traditional processor datapath, especially for reconfigurable platforms such as FPGAs. This paper presents an architecture integrating NoC concepts into the design of a processor datapath in order to create a polymorphic processor. The paper will further explore and analyze the effect of topology choices and NoC design on the performance of polymorphic processors with specific focus on the impacts of NoC datapath integration.

About this research paper

What this paper is about

Polymorphic processing has the goal of producing a processor combining the advantages of general purpose processing with the significant gains achievable by custom application specific computing. To achieve these ends a novel polymorphic processor architecture is presented. Incorporating networking on a chip (NoC) techniques into the datapath design has the potential to provide noticeable advantages when compared to a traditional processor datapath, especially for reconfigurable platforms such as FPGAs. This paper presents an architecture integrating NoC concepts into the design of a processor datapath in order to create a polymorphic processor. The paper will further explore and analyze the effect of topology choices and NoC design on the performance of polymorphic processors with specific focus on the impacts of NoC datapath integration.

Why it matters

OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

Polymorphic processing has the goal of producing a processor combining the advantages of general purpose processing with the significant gains achievable by custom application specific computing. To achieve these ends a novel polymorphic processor architecture is presented. Incorporating networking on a chip (NoC) techniques into the datapath design has the potential to provide noticeable advantages when compared to a traditional processor datapath, especially for reconfigurable platforms such as FPGAs. This paper presents an architecture integrating NoC concepts into the design of a processor datapath in order to create a polymorphic processor. The paper will further explore and analyze the effect of topology choices and NoC design on the performance of polymorphic processors with specific focus on the impacts of NoC datapath integration.

Key concepts: Datapath, Computer architecture, Computer science, Field-programmable gate array, Embedded system, Microarchitecture, Parallel computing, Reconfigurable computing

Related papers

Back to paper searchBrowse research topicsOriginal source
NoC datapath for polymorphic processors in embedded systems — Research Paper | ScholarLens