Performance enhancement and Energy reduction in multi core Architecture with network on chip Interconnection
Loading...
Date
item.page.authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
With the rapid advancements in the semiconductor industry and
newlinevarious VLSI-based applications, multi-core technology has evolved as a
newlinehighly dominant field. A multi-core processor comprises numerous cores on a
newlinesingle die. A multi-core processor supports parallel computation by
newlineperforming thread-level parallelism thereby achieving faster execution of
newlinetasks. A multi-core processor technology is widely used across many
newlineapplication domains due to its improved speed and performance and
newlinesupporting high-speed computing devices. The interconnectivity between the
newlinecores of a multi-core architecture becomes a bottleneck with the bus-based
newlinesystems. The Network-on-Chip (NoC) paradigm comes into the picture to
newlineovercome the drawbacks involved in bus-based connectivity. The Intellectual
newlineProperty (IP) cores of a multi-core architecture are connected through the
newlinelinks, routers, and the Network Interface (NI).
newlineThere are two kinds of multi-core processors namely homogeneous
newlineand heterogeneous configurations. This classification is based on the type and
newlinesize of the cores. A homogenous category has identical cores, and they are
newlinereferred to as regular topologies. The salient features of a regular topology are
newlineits fault tolerance and reusability. In the heterogeneous configuration also
newlinecalled the irregular topology, the cores are of different sizes, and they are
newlinedesigned for a particular application and hence they are referred to as
newlineapplication specific topologies.
newlineThere are many benefits that come with the multi-core processors
newlinelike high speed and parallel computation thereby an improved performance
newlinealong with the reduced time taken for the execution of the tasks
newline