17688288. NEURAL NETWORK OPERATION METHOD AND APPARATUS simplified abstract (Samsung Electronics Co., Ltd.)

NEURAL NETWORK OPERATION METHOD AND APPARATUS

Organization Name

Samsung Electronics Co., Ltd.

Inventor(s)

Maksim Ostapenko of Suwon-si (KR)

Hanwoong Jung of Seoul (KR)

NEURAL NETWORK OPERATION METHOD AND APPARATUS - A simplified explanation of the abstract

This abstract first appeared for US patent application 17688288 titled 'NEURAL NETWORK OPERATION METHOD AND APPARATUS

Simplified Explanation

The patent application describes a method and apparatus for operating a neural network. Here is a simplified explanation of the abstract:

• The method involves receiving data for a neural network operation.
• It determines whether the size of the data is less than or equal to a threshold.
• If the size is within the threshold, it generates stacked data by stacking a portion of the data.
• Finally, it performs the neural network operation in parallel based on the stacked data.

Potential applications of this technology:

• This method can be used in various fields where neural networks are employed, such as image recognition, natural language processing, and data analysis.
• It can enhance the efficiency and speed of neural network operations, making it suitable for real-time applications.

Problems solved by this technology:

• Neural network operations can be computationally intensive, especially with large datasets. This method helps optimize the process by parallelizing the operation and reducing the computational load.
• It addresses the challenge of handling large datasets by stacking a portion of the data, allowing for efficient processing without compromising accuracy.

Benefits of this technology:

• Improved efficiency: By performing neural network operations in parallel, the method reduces the overall processing time, enabling faster results.
• Scalability: The method can handle large datasets by stacking data, making it suitable for applications with extensive data requirements.
• Enhanced accuracy: Despite stacking data, the method ensures accuracy by considering the portion of data that meets the threshold, minimizing any potential loss of information.

Original Abstract Submitted

A neural network operation method and apparatus are disclosed, where the network operation method including receiving data for a neural network operation, determining whether a size of the data is less than or equal to a threshold, generating stacked data by stacking a portion of the data based on the determining, and performing the neural network operation in parallel based on the stacked data.