DocumentCode :
2541196
Title :
IC design of 8×8 digital CNN with optoelectronic interface
Author :
Jankowsk, S. ; Wielgus, A. ; Pleskacz, Witold A. ; Buczynski, R. ; Wisniewski, M.
Author_Institution :
Inst. of Electron. Syst., Warsaw Univ. of Technol., Poland
fYear :
2000
fDate :
2000
Firstpage :
431
Lastpage :
436
Abstract :
Presents an optoelectronic CNN system which consists of two integrated circuits connected through flip-chip bonding: a) the photonic GaAs matrix of receivers and emitter; b) the CMOS chip of digital CNN cells with programmable 5-bit weights and threshold currents. This solution is easier to design and is more reliable than analogue circuits. The main advantages of our system include parallel input/output of considered image and high precision of programmable weights. Although the processing speed of the digital CNN is considerably lower than that of analogue implementations, it is fast enough for efficient image processing by complex CNN programs
Keywords :
CMOS digital integrated circuits; cellular neural nets; circuit CAD; flip-chip devices; image processing; integrated circuit design; integrated optoelectronics; neural chips; neural net architecture; 8×8 digital CNN; CMOS chip; flip-chip bonding; optoelectronic interface; parallel input/output; photonic GaAs matrix; programmable 5-bit weights; threshold currents; Bonding; CMOS digital integrated circuits; CMOS integrated circuits; Cellular neural networks; Digital integrated circuits; Gallium arsenide; Integrated circuit reliability; Optical receivers; Photonic integrated circuits; Threshold current;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Cellular Neural Networks and Their Applications, 2000. (CNNA 2000). Proceedings of the 2000 6th IEEE International Workshop on
Conference_Location :
Catania
Print_ISBN :
0-7803-6344-2
Type :
conf
DOI :
10.1109/CNNA.2000.877367
Filename :
877367
Link To Document :
بازگشت