Article
Refine
Year of publication
- 2024 (30)
- 2023 (64)
- 2022 (78)
- 2021 (86)
- 2020 (101)
- 2019 (95)
- 2018 (84)
- 2017 (72)
- 2016 (78)
- 2015 (83)
- 2014 (93)
- 2013 (97)
- 2012 (82)
- 2011 (130)
- 2010 (121)
- 2009 (121)
- 2008 (103)
- 2007 (94)
- 2006 (86)
- 2005 (99)
- 2004 (131)
- 2003 (74)
- 2002 (92)
- 2001 (88)
- 2000 (84)
- 1999 (88)
- 1998 (82)
- 1997 (79)
- 1996 (70)
- 1995 (68)
- 1994 (76)
- 1993 (51)
- 1992 (48)
- 1991 (25)
- 1990 (35)
- 1989 (38)
- 1988 (54)
- 1987 (32)
- 1986 (18)
- 1985 (32)
- 1984 (18)
- 1983 (17)
- 1982 (26)
- 1981 (18)
- 1980 (35)
- 1979 (23)
- 1978 (30)
- 1977 (14)
- 1976 (13)
- 1975 (10)
- 1974 (3)
- 1972 (2)
- 1971 (1)
- 1968 (1)
Institute
- Fachbereich Medizintechnik und Technomathematik (1356)
- INB - Institut für Nano- und Biotechnologien (503)
- Fachbereich Chemie und Biotechnologie (472)
- Fachbereich Elektrotechnik und Informationstechnik (413)
- IfB - Institut für Bioengineering (409)
- Fachbereich Energietechnik (361)
- Fachbereich Luft- und Raumfahrttechnik (253)
- Fachbereich Maschinenbau und Mechatronik (147)
- Fachbereich Wirtschaftswissenschaften (116)
- Fachbereich Bauingenieurwesen (69)
Language
- English (3273) (remove)
Document Type
- Article (3273) (remove)
Keywords
- Einspielen <Werkstoff> (7)
- avalanche (5)
- Earthquake (4)
- FEM (4)
- Finite-Elemente-Methode (4)
- LAPS (4)
- biosensors (4)
- field-effect sensor (4)
- frequency mixing magnetic detection (4)
- CellDrum (3)
This paper describes the realization of a novel neurocomputer which is based on the concepts of a coprocessor. In contrast to existing neurocomputers the main interest was the realization of a scalable, flexible system, which is capable of computing neural networks of arbitrary topology and scale, with full independence of special hardware from the software's point of view. On the other hand, computational power should be added, whenever needed and flexibly adapted to the requirements of the application. Hardware independence is achieved by a run time system which is capable of using all available computing power, including multiple host CPUs and an arbitrary number of neural coprocessors autonomously. The realization of arbitrary neural topologies is provided through the implementation of the elementary operations which can be found in most neural topologies.