Tampere University of Technology

TUTCRIS Research Portal

Machine learning: How it can help nanocomputing

Research output: Contribution to journalReview ArticleScientificpeer-review


Original languageEnglish
Pages (from-to)1347-1363
Number of pages17
JournalJournal of Computational and Theoretical Nanoscience
Issue number8
Publication statusPublished - Aug 2011
Publication typeA2 Review article in a scientific journal


We examine possibilities for making advances in nanocomputing by bringing in ideas from the field of machine learning. The potential from combining machine learning with nanocomputing seems to be underutilized. We review three complementary approaches. Firstly, machine learning can be used in the different phases of developing complicated nanocomputing devices: in modeling, designing, constructing, and programming the devices. Secondly, machine learning methods implemented by nanocomputing hardware can be a competitive solution especially for specialized application areas like sensory information processing; working towards such implementations advances nanocomputing by guiding development of the nanocomponents and architectures required for such applications. Thirdly, nanotechnology enabled quantum computing can significantly increase our capacity to solve NP-complete optimization problems; although this increase is not specific to machine learning, several such problems occur in machine learning and artificial intelligence, hence solving such problems is a useful goal that partly motivates development of quantum computing. The main value of this paper is to provide new ideas for researchers working on nanocomputing, nanoarchitectures, development and design of nanoprocessors and other nanocomponents, or nanomanufacturing.


  • Machine Learning, Nanoarchitectures, Nanocomputing, Nanotechnology, Neural Networks, Quantum Computing