By Yoonjin Kim

Coarse-grained reconfigurable structure (CGRA) has emerged as an answer for versatile, application-specific optimization of embedded platforms. supporting you realize the problems excited by designing and developing embedded platforms, layout of Low-Power Coarse-Grained Reconfigurable Architectures bargains new frameworks for optimizing the structure of parts in embedded platforms so that it will reduce sector and store energy. genuine software benchmarks and gate-level simulations substantiate those frameworks. the 1st 1/2 the ebook explains how one can lessen energy within the configuration cache. The authors current a low-power reconfiguration process in response to reusable context pipelining that merges the idea that of context reuse into context pipelining. in addition they suggest dynamic context compression able to helping required bits of the context phrases set to let and the redundant bits set to disable. moreover, they talk about dynamic context administration for lowering energy intake within the configuration cache by means of controlling a read/write operation of the redundant context phrases. targeting the layout of a cheap processing aspect array to lessen quarter and tool intake, the second one 1/2 the textual content offers a cheap array textile that uniquely rearranges processing parts and their interconnection designs. The ebook additionally describes hierarchical reconfigurable computing arrays such as reconfigurable computing blocks with forms of communique constitution. the 2 computing blocks percentage serious assets, providing a good verbal exchange interface among them and lowering the final zone. the ultimate bankruptcy takes an built-in method of optimization that pulls at the layout schemes provided in previous chapters. utilizing a case research, the authors display the synergy influence of mixing a number of layout schemes.

Show description

Read or Download Design of Low-Power Coarse-Grained Reconfigurable Architectures PDF

Similar electrical & electronics books

MIMO System Technology for Wireless Communications (Electrical Engineering & Applied Signal Processing Series)

For broadband communications, it was once frequency department multiplexing. For optical communications, it was once wavelength department multiplexing. Then, for all sorts of networks it used to be code department. Breakthroughs in transmission pace have been made attainable via those advancements, heralding next-generation networks of accelerating power in every one case.

Sensor Network Operations

This wonderful name introduces the idea that of mission-oriented sensor networks as disbursed dynamic structures of interacting sensing units which are networked to together execute advanced real-time missions lower than uncertainity. It presents the newest, but unpublished effects at the major technical and alertness demanding situations of mission-oriented sensor networks.

Extra resources for Design of Low-Power Coarse-Grained Reconfigurable Architectures

Example text

The MIPS ISA is extended for the REMARC using special instructions. The main processor issues these instructions to the REMARC which executes them in a manner similar to a floating point coprocessor. 7: Block diagram of a microprocessor with REMARC. (From T. Miyamori and K. ” In Proceedings of IEEE Symposium on FPGAs for Custom Computing Machines, c 1998 IEEE. ) © 2011 by Taylor and Francis Group, LLC Trends in CGRA 15 the functions of reconfigurable coprocessor instructions are configurable (or programmable) so that they can be specialized for specific applications.

The methodology was to start with standard cells. Placement was done manually and routing was partially automatic, partially manual. The layout implementation is based on standard cells. Medium drive strength is used inside a subblock and strong buffering for the lines, which broadcast inside the RC. The cells mostly are hand-placed, because many parts of the microarchitecture are regular. Routing inside a subblock is done automatically. The routing between the subblocks is done manually. 196 mm2 .

First of all, we describe entire structure, PE array fabric and memory organization of the CGRA examples in the architecture aspect. In the design space exploration, we introduce architecture optimization issues and existing DSEs to generate efficient CGRA structures. 25: PipeRench PE floorplan. (From H. Schmit, D. Whelihan, M. Moe, A. Tsai, B. Levine, and R. 18 micron technology,” In Proceedings of IEEE Custom Integrated Circuits Conference, c 2002 IEEE. ) of existing CGRAs. Finally, we illustrate physical implementation examples of CGRAs to show analysis of their hardware costs.

Download PDF sample

Rated 4.25 of 5 – based on 43 votes