Cambricon Technologies has been granted a patent for an apparatus that facilitates backpropagation of a multilayer neural network (MNN). The invention involves a computation circuit that processes MNN data using discrete values and bitwise operations, enhancing the efficiency of neural network processing. GlobalData’s report on Cambricon Technologies gives a 360-degree view of the company including its patenting strategy. Buy the report here.
Smarter leaders trust GlobalData
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
According to GlobalData’s company profile on Cambricon Technologies, User journey analytics was a key innovation area identified from patents. Cambricon Technologies's grant share as of May 2024 was 36%. Grant share is based on the ratio of number of grants to total number of patents.
Neural network processor for backpropagation with discrete values
A recently granted patent (Publication Number: US11995554B2) discloses an innovative apparatus for backpropagation of a multilayer neural network (MNN). The apparatus includes a computation circuit with a master computation circuit and one or more slave computation circuits. The master computation circuit calculates an input gradient vector based on a first output gradient vector and performs bitwise operations on discrete values stored in a lookup table. The slave computation circuits parallelly calculate portions of a second output vector. An interconnection circuit combines these portions to generate the second output gradient vector. Additionally, the apparatus includes a controller circuit to initiate the backpropagation process and manage instructions for computation.
Furthermore, the patent details a method for backpropagation of MNN using the described apparatus. The method involves receiving MNN data, calculating input gradient vectors, and parallelly computing output vectors. The interconnection circuit facilitates combining these vectors. The master computation circuit handles bitwise operations and activation function derivatives, while the slave computation circuits manage weight value gradients and updates. The method also includes data conversion processes to handle continuous and discrete data types efficiently. Overall, this patent introduces a comprehensive approach to backpropagation in MNNs, enhancing computational efficiency and accuracy through parallel processing and optimized data handling mechanisms.
To know more about GlobalData’s detailed insights on Cambricon Technologies, buy the report here.
Data Insights
From
The gold standard of business intelligence.
Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.
GlobalData, the leading provider of industry intelligence, provided the underlying data, research, and analysis used to produce this article.
GlobalData Patent Analytics tracks bibliographic data, legal events data, point in time patent ownerships, and backward and forward citations from global patenting offices. Textual analysis and official patent classifications are used to group patents into key thematic areas and link them to specific companies across the world’s largest industries.
