Title :
Ex-situ programming in a neuromorphic memristor based crossbar circuit
Author :
Chris Yakopcic;Tarek M. Taha
Author_Institution :
Department of Electrical and Computer Engineering, University of Dayton, Dayton, OH, USA
fDate :
6/1/2015 12:00:00 AM
Abstract :
This paper discusses a feedback programming method for a high density crossbar. This programming technique is capable of operating without the use of any transistor or diode isolation at the memristor crosspoints. A series of reads is applied to the crossbar before each write that is able to determine the resistance of each memristor in the crossbar despite the many parallel resistance paths. This is essential because the variation observed in memristor crossbars makes programming very difficult when using just a single write pulse without error checking. The programming method is then used to program a neuromorphic crossbar. Results show successful ex-situ training of a high density crossbar with significant area savings when compared to a one transistor one memristor (1T1M) design. A comparison between different crossbar designs is performed relative to the A-to-D complexity required to program each circuit for a varying device resistance ratio and programming precision.
Keywords :
"Memristors","Resistance","Programming","Mathematical model","Integrated circuit modeling","SPICE","Transistors"
Conference_Titel :
Aerospace and Electronics Conference (NAECON), 2015 National
Electronic_ISBN :
2379-2027
DOI :
10.1109/NAECON.2015.7443087