We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.DC

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Distributed, Parallel, and Cluster Computing

Title: High Performance and Optimal Configuration of Accurate Heterogeneous Block-Based Approximate Adder

Abstract: Approximate computing is an emerging paradigm to improve power and performance efficiency for error-resilient application. Recent approximate adders have significantly extended the design space of accuracy-power configurable approximate adders, and find optimal designs by exploring the design space. In this paper, a new energy-efficient heterogeneous block-based approximate adder (HBBA) is proposed; which is a generic/configurable model that can be transformed to a particular adder by defining some configurations. An HBBA, in general, is composed of heterogeneous sub-adders, where each sub-adder can have a different configuration. A set of configurations of all the sub-adders in an HBBA defines its configuration. The block-based adders are approximated through inexact logic configuration and truncated carry chains. HBBA increases design space providing additional design points that fall on the Pareto-front and offer better power-accuracy trade-off compared to other configurations. Furthermore, to avoid Mont-Carlo simulations, we propose an analytical modelling technique to evaluate the probability of error and Probability Mass Function (PMF) of error value. Moreover, the estimation method estimates delay, area and power of heterogeneous block-based approximate adders. Thus, based on the analytical model and estimation method, the optimal configuration under a given error constraint can be selected from the whole design space of the proposed adder model by exhaustive search. The simulation results show that our HBBA provides improved accuracy in terms of error metrics compared to some state-of-the-art approximate adders. HBBA with 32 bits length serves about 15% reduction in area and up to 17% reduction in energy compared to state-of-the-art approximate adders.
Comments: Submitted to the IEEE-TCAD journal, 16 pages, 16 figures
Subjects: Distributed, Parallel, and Cluster Computing (cs.DC); Hardware Architecture (cs.AR)
Cite as: arXiv:2106.08800 [cs.DC]
  (or arXiv:2106.08800v1 [cs.DC] for this version)

Submission history

From: Ebrahim Farahmand [view email]
[v1] Wed, 16 Jun 2021 14:03:49 GMT (4584kb,D)

Link back to: arXiv, form interface, contact.