Configurable neural network integrated circuit

Details Configurable neural network integrated circuit Configurable neural network integrated circuit

1995-07-28

1999-09-21

MacDonald, Allen R.

Data processing: artificial intelligence

Neural network

Structure

395 24, G06F 1518

Patent

active

059567031

ABSTRACT:
A neural network IC 31 includes n dedicated processing elements (PEs) 62, an output register 66 for storing the PEs' outputs so that they are immediately accessible to all of the PEs, a number of output circuits 78 that are connected to selected PEs to provide binary outputs, and a timing circuit 74. Each of the PEs includes a weight memory 90 for storing input, output and bias weight arrays, a first in first out (FIFO) memory 88 for storing input data, a dot product circuit 92 and an activation circuit 94. The dot product circuit computes a dot product of the input weight array and the contents of the FIFO memory, a dot product of the output weight array and the contents of the output register, a dot product of the bias value and a constant, and sums the three results. The activation circuit maps the output of the dot product circuit through an activation function to produce the PE's output. The inclusion of a memory 90 that stores both input and output weight arrays in conjunction with the output register 66 allows the PEs to be configured to implement arbitrary feed-forward and recurrent neural network architectures.

REFERENCES:
patent: 5073867 (1991-12-01), Murphy et al.
patent: 5142666 (1992-08-01), Yoshizawa et al.
patent: 5165010 (1992-11-01), Masuda et al.
patent: 5170463 (1992-12-01), Fujimoto et al.
patent: 5216751 (1993-06-01), Gardner et al.
patent: 5278945 (1994-01-01), Basehore et al.
patent: 5303311 (1994-04-01), Epting et al.
patent: 5339242 (1994-08-01), Reid et al.
patent: 5384896 (1995-01-01), Sakaue et al.
patent: 5524175 (1996-06-01), Sato et al.
Chang et al., Design of multiprocessor DSP chip for flexible information processing, ICASSP-92, pp. 637-640, Mar. 26, 1992.
Chang et al., Digital VLSI multi-processor design for neuro-computers, IJCNN-92, pp. 1-6, Jun. 11, 1992.
Aihara et al., A sparse memory-access neural network engine with 96 parallel data-driven processing units., IEEE-95, pp. 72-73, Feb. 17, 1995.
A VLSI Solution to On-chip Learning Neural Networks, Adaptive Solutions, Inc., Beaverton OR.
Association Engine User's Manual, Dec. 1993, Motorola.
Electrically Trainable Analog Neural Network, 80170NX, Intel Corporation, 1991, pp. 2-15.
High-Density Complementary Metal Oxide Semiconductor (HCMOS) Microcontroller Unit, HC11, MC68HC11F1, Technical Data, (Rev.2)Motorola, Inc., 1990.

Affiliated with

Also associated with

Rating

Configurable neural network integrated circuit does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Configurable neural network integrated circuit, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Configurable neural network integrated circuit will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-91625