{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/autoax-an-automatic-design-space-exploration","title":"autoAx: An Automatic Design Space Exploration and Circuit Building Methodology utilizing Libraries of Approximate Components","arxiv_id":"1902.10807","date":"2019-02-22","proceeding":null,"authors":["Vojtech Mrazek","Muhammad Abdullah Hanif","Zdenek Vasicek","Lukas Sekanina","Muhammad Shafique"],"abstract":"Approximate computing is an emerging paradigm for developing highly\nenergy-efficient computing systems such as various accelerators. In the\nliterature, many libraries of elementary approximate circuits have already been\nproposed to simplify the design process of approximate accelerators. Because\nthese libraries contain from tens to thousands of approximate implementations\nfor a single arithmetic operation it is intractable to find an optimal\ncombination of approximate circuits in the library even for an application\nconsisting of a few operations. An open problem is \"how to effectively combine\ncircuits from these libraries to construct complex approximate accelerators\".\nThis paper proposes a novel methodology for searching, selecting and combining\nthe most suitable approximate circuits from a set of available libraries to\ngenerate an approximate accelerator for a given application. To enable fast\ndesign space generation and exploration, the methodology utilizes machine\nlearning techniques to create computational models estimating the overall\nquality of processing and hardware cost without performing full synthesis at\nthe accelerator level. Using the methodology, we construct hundreds of\napproximate accelerators (for a Sobel edge detector) showing different but\nrelevant tradeoffs between the quality of processing and hardware cost and\nidentify a corresponding Pareto-frontier. Furthermore, when searching for\napproximate implementations of a generic Gaussian filter consisting of 17\narithmetic operations, the proposed approach allows us to identify\napproximately $10^3$ highly important implementations from $10^{23}$ possible\nsolutions in a few hours, while the exhaustive search would take four months on\na high-end processor.","url_abs":"http://arxiv.org/abs/1902.10807v2","url_pdf":"http://arxiv.org/pdf/1902.10807v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"autoax-an-automatic-design-space-exploration","repo_url":"https://github.com/ehw-fit/autoxfpgas","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null},{"paper_slug":"autoax-an-automatic-design-space-exploration","repo_url":"https://github.com/ehw-fit/xel-fpgas","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}