{"291281":{"#nid":"291281","#data":{"type":"news","title":"Neuromorphic Computing \u0022Roadmap\u0022 Envisions Analog Path to Simulating Human Brain","body":[{"value":"\u003Cp\u003EIn the field of neuromorphic engineering, researchers study computing techniques that could someday mimic human cognition. Electrical engineers at the Georgia Institute of Technology recently published a \u0022roadmap\u0022 that details innovative analog-based techniques that could make it possible to build a practical neuromorphic computer.\u003C\/p\u003E\u003Cp\u003EA core technological hurdle in this field involves the electrical power requirements of computing hardware. Although a human brain functions on a mere 20 watts of electrical energy, a digital computer that could approximate human cognitive abilities would require tens of thousands of integrated circuits (chips) and a hundred thousand watts of electricity or more \u2013 levels that exceed practical limits.\u003C\/p\u003E\u003Cp\u003EThe Georgia Tech roadmap proposes a solution based on analog computing techniques, which require far less electrical power than traditional digital computing. The more efficient analog approach would help solve the daunting cooling and cost problems that presently make digital neuromorphic hardware systems impractical.\u0026nbsp;\u003C\/p\u003E\u003Cp\u003E\u0022To simulate the human brain, the eventual goal would be large-scale neuromorphic systems that could offer a great deal of computational power, robustness and performance,\u0022 said \u003Ca href=\u0022http:\/\/www.ece.gatech.edu\/faculty-staff\/fac_profiles\/bio.php?id=45\u0022\u003EJennifer Hasler\u003C\/a\u003E, a professor in the Georgia Tech \u003Ca href=\u0022http:\/\/www.ece.gatech.edu\/\u0022\u003ESchool of Electrical and Computer Engineering\u003C\/a\u003E (ECE), who is a pioneer in using analog techniques for neuromorphic computing. \u0022A configurable analog-digital system can be expected to have a power efficiency improvement of up to 10,000 times compared to an all-digital system.\u0022\u003C\/p\u003E\u003Cp\u003EHasler and a former student recently published a detailed plan that describes the development of computer systems capable of human-like cognition. The paper, \u0022Finding a Roadmap to Achieve Large Neuromorphic Hardware Systems\u0022 by Hasler and Bo Marr, was published in the September 2013 edition of the journal \u003Cem\u003EFrontiers in Neuroscience\u003C\/em\u003E.\u003C\/p\u003E\u003Cp\u003E\u0022To my knowledge, this is the first time a detailed neuromorphic roadmap has been attempted,\u0022 said Hasler. \u0022We describe specific computational techniques could offer real progress in neuromorphic systems.\u0022\u003C\/p\u003E\u003Cp\u003EUnlike digital computing, in which computers can address many different applications by processing different software programs, analog circuits have traditionally been hard-wired to address a single application. For example, cell phones use energy-efficient analog circuits for a number of specific functions, including capturing the user\u0027s voice, amplifying incoming voice signals, and controlling battery power.\u003C\/p\u003E\u003Cp\u003EBecause analog devices do not have to process binary codes as digital computers do, their performance can be both faster and much less power hungry. Yet traditional analog circuits are limited because they\u0027re built for a specific application, such as processing signals or controlling power. They don\u0027t have the flexibility of digital devices that can process software, and they\u0027re vulnerable to signal disturbance issues, or noise.\u003C\/p\u003E\u003Cp\u003EIn recent years, Hasler has developed a new approach to analog computing, in which silicon-based analog integrated circuits take over many of the functions now performed by familiar digital integrated circuits. These analog chips can be quickly reconfigured to provide a range of processing capabilities, in a manner that resembles conventional digital techniques in some ways.\u003C\/p\u003E\u003Cp\u003EOver the last several years, Hasler and her research group have developed devices called field programmable analog arrays (FPAA). Like field programmable gate arrays (FPGA), which are digital integrated circuits that are ubiquitous in modern computing, the FPAA can be reconfigured after it\u0027s manufactured \u2013 hence the phrase \u0022field-programmable.\u0022\u003C\/p\u003E\u003Cp\u003EHasler and Marr\u0027s 29-page paper traces a development process that could lead to the goal of reproducing human-brain complexity. The researchers investigate in detail a number of intermediate steps that would build on one another, helping researchers advance the technology sequentially.\u003C\/p\u003E\u003Cp\u003EFor example, the researchers discuss ways to scale energy efficiency, performance and size in order to eventually achieve large-scale neuromorphic systems. The authors also address how the implementation and the application space of neuromorphic systems can be expected to evolve over time.\u003C\/p\u003E\u003Cp\u003E\u0022A major concept here is that we have to first build smaller systems capable of a simple representation of one layer of human brain cortex,\u0022 Hasler said. \u0022When that system has been successfully demonstrated, we can then replicate it in ways that increase its complexity and performance.\u0022\u003C\/p\u003E\u003Cp\u003EAmong neuromorphic computing\u0027s major hurdles are the communication issues involved in networking integrated circuits in ways that could replicate human cognition. In their paper, Hasler and Marr emphasize local interconnectivity to reduce complexity. Moreover, they argue it\u0027s possible to achieve these capabilities via purely silicon-based techniques, without relying on novel devices that are based on other approaches.\u003C\/p\u003E\u003Cp\u003ECommenting on the recent publication, Alice C. Parker, a professor of electrical engineering at the University of Southern California, said, \u0022Professor Hasler\u0027s technology roadmap is the first deep analysis of the prospects for large scale neuromorphic intelligent systems, clearly providing practical guidance for such systems, with a nearer-term perspective than our whole-brain emulation predictions. Her expertise in analog circuits, technology and device models positions her to provide this unique perspective on neuromorphic circuits.\u0022\u0026nbsp; \u0026nbsp;\u003C\/p\u003E\u003Cp\u003EEugenio Culurciello, an associate professor of biomedical engineering at Purdue University, commented, \u0022I find this paper to be a very accurate description of the field of neuromorphic data processing systems. Hasler\u0027s devices provide some of the best performance per unit power I have ever seen and are surely on the roadmap for one of the major technologies of the future.\u0022\u003C\/p\u003E\u003Cp\u003ESaid Hasler: \u0022In this study, we conclude that useful neural computation machines based on biological principles \u2013 and potentially at the size of the human brain -- seems technically within our grasp. We think that it\u0027s more a question of gathering the right research teams and finding the funding for research and development than of any insurmountable technical barriers.\u0022\u003Cbr \/\u003E\u003Cbr \/\u003E\u003Cstrong\u003EResearch News\u003C\/strong\u003E\u003Cbr \/\u003E\u003Cstrong\u003EGeorgia Institute of Technology\u003C\/strong\u003E\u003Cbr \/\u003E\u003Cstrong\u003E177 North Avenue\u003C\/strong\u003E\u003Cbr \/\u003E\u003Cstrong\u003EAtlanta, Georgia\u0026nbsp; 30332-0181\u0026nbsp; USA\u003C\/strong\u003E\u003Cbr \/\u003E\u003Cbr \/\u003E\u003Cstrong\u003EMedia Relations Contacts\u003C\/strong\u003E: John Toon (\u003Ca href=\u0022mailto:jtoon@gatech.edu\u0022\u003Ejtoon@gatech.edu\u003C\/a\u003E) (404-894-6986) or Brett Israel (\u003Ca href=\u0022mailto:brett.israel@comm.gatech.edu\u0022\u003Ebrett.israel@comm.gatech.edu\u003C\/a\u003E) (404-385-1933).\u003C\/p\u003E\u003Cp\u003E\u003Cstrong\u003EWriter\u003C\/strong\u003E: Rick Robinson\u003Cbr \/\u003E\u003Cbr \/\u003E\u003C\/p\u003E","summary":null,"format":"limited_html"}],"field_subtitle":"","field_summary":[{"value":"\u003Cp\u003EIn the field of neuromorphic engineering, researchers study computing techniques that could someday mimic human cognition. Electrical engineers at the Georgia Institute of Technology recently published a \u0022roadmap\u0022 that details innovative analog-based techniques that could make it possible to build a practical neuromorphic computer.\u003C\/p\u003E","format":"limited_html"}],"field_summary_sentence":[{"value":"Georgia Tech researchers have published a \u0022roadmap\u0022 that details techniques that could make it possible to build a practical neuromorphic computer to mimic human cognition."}],"uid":"27303","created_gmt":"2014-04-16 16:16:20","changed_gmt":"2016-10-08 03:16:15","author":"John Toon","boilerplate_text":"","field_publication":"","field_article_url":"","dateline":{"date":"2014-04-16T00:00:00-04:00","iso_date":"2014-04-16T00:00:00-04:00","tz":"America\/New_York"},"extras":[],"hg_media":{"291251":{"id":"291251","type":"image","title":"Neuromorphic computing3","body":null,"created":"1449244289","gmt_created":"2015-12-04 15:51:29","changed":"1475894988","gmt_changed":"2016-10-08 02:49:48","alt":"Neuromorphic computing3","file":{"fid":"199241","name":"14c10202-p10-005a.jpg","image_path":"\/sites\/default\/files\/images\/14c10202-p10-005a_0.jpg","image_full_path":"http:\/\/www.tlwarc.hg.gatech.edu\/\/sites\/default\/files\/images\/14c10202-p10-005a_0.jpg","mime":"image\/jpeg","size":1174987,"path_740":"http:\/\/www.tlwarc.hg.gatech.edu\/sites\/default\/files\/styles\/740xx_scale\/public\/images\/14c10202-p10-005a_0.jpg?itok=y5f7leLO"}},"291241":{"id":"291241","type":"image","title":"Neuromorphic computing2","body":null,"created":"1449244289","gmt_created":"2015-12-04 15:51:29","changed":"1475894988","gmt_changed":"2016-10-08 02:49:48","alt":"Neuromorphic computing2","file":{"fid":"199240","name":"14c10202-p10-003a.jpg","image_path":"\/sites\/default\/files\/images\/14c10202-p10-003a_0.jpg","image_full_path":"http:\/\/www.tlwarc.hg.gatech.edu\/\/sites\/default\/files\/images\/14c10202-p10-003a_0.jpg","mime":"image\/jpeg","size":1501298,"path_740":"http:\/\/www.tlwarc.hg.gatech.edu\/sites\/default\/files\/styles\/740xx_scale\/public\/images\/14c10202-p10-003a_0.jpg?itok=YYZJUxKo"}},"291231":{"id":"291231","type":"image","title":"Neuromorphic computing","body":null,"created":"1449244289","gmt_created":"2015-12-04 15:51:29","changed":"1475894988","gmt_changed":"2016-10-08 02:49:48","alt":"Neuromorphic computing","file":{"fid":"199239","name":"14c10202-p10-001a.jpg","image_path":"\/sites\/default\/files\/images\/14c10202-p10-001a_0.jpg","image_full_path":"http:\/\/www.tlwarc.hg.gatech.edu\/\/sites\/default\/files\/images\/14c10202-p10-001a_0.jpg","mime":"image\/jpeg","size":1080937,"path_740":"http:\/\/www.tlwarc.hg.gatech.edu\/sites\/default\/files\/styles\/740xx_scale\/public\/images\/14c10202-p10-001a_0.jpg?itok=-y_T4dRN"}}},"media_ids":["291251","291241","291231"],"groups":[{"id":"1188","name":"Research Horizons"}],"categories":[{"id":"145","name":"Engineering"},{"id":"146","name":"Life Sciences and Biology"},{"id":"135","name":"Research"}],"keywords":[{"id":"7569","name":"analog"},{"id":"1912","name":"brain"},{"id":"91641","name":"human cognition"},{"id":"91651","name":"Jennifer Hasler"},{"id":"91631","name":"neuromorphic computing"},{"id":"166855","name":"School of Electrical and Computer Engineering"}],"core_research_areas":[{"id":"39451","name":"Electronics and Nanotechnology"},{"id":"39481","name":"National Security"}],"news_room_topics":[{"id":"71881","name":"Science and Technology"}],"event_categories":[],"invited_audience":[],"affiliations":[],"classification":[],"areas_of_expertise":[],"news_and_recent_appearances":[],"phone":[],"contact":[{"value":"\u003Cp\u003EJohn Toon\u003C\/p\u003E\u003Cp\u003EResearch News\u003C\/p\u003E\u003Cp\u003E\u003Ca href=\u0022mailto:jtoon@gatech.edu\u0022\u003Ejtoon@gatech.edu\u003C\/a\u003E\u003C\/p\u003E\u003Cp\u003E(404) 894-6986\u003C\/p\u003E","format":"limited_html"}],"email":["jtoon@gatech.edu"],"slides":[],"orientation":[],"userdata":""}}}