Neural network hardware circuitry and data selection methods, chips, and electronic devices

TW202642407APending Publication Date: 2026-10-16VERISILICON TECH (SHANGHAI) CO LTD +1
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
TW114144194
Authority / Receiving Office
TW · TW
Patent Type
Applications
Current Assignee / Owner
Priority Date
2025-03-31
Filing Date
2025-11-13
Publication Date
2026-10-16

Smart Images

  • Figure 00000000_0000_ABST
    Figure 00000000_0000_ABST
Patent Text Reader

Abstract

This application relates to a hardware circuit for neural networks and a data selection method, a chip, and an electronic device, belonging to the field of neural networks. The hardware circuit for neural networks includes: an input data multiplexer, an actuator, and an output data multiplexer, with the actuator located between the input data multiplexer and the output data multiplexer. The input data multiplexer is used to select external input data or internal data of the hardware circuit for use as the input data source according to a first selection control signal. The actuator includes at least one layer of corresponding custom hardware circuitry in the neural network. The output data multiplexer is used to select whether to output the output data of the actuator as external data or as the internal data according to a second selection control signal. This application can reduce the circuit area and power consumption during the hardware implementation of neural networks.
Need to check novelty before this filing date? Find Prior Art