International business machines corporation (20240289607). CO-DESIGN OF A MODEL AND CHIP FOR DEEP LEARNING BACKGROUND simplified abstract

CO-DESIGN OF A MODEL AND CHIP FOR DEEP LEARNING BACKGROUND

Organization Name

international business machines corporation

Inventor(s)

Irem Boybat Kara of Adliswil (CH)

Hadjer Benmeziane of AULNOY-LEZ-VALENCIENNES (FR)

Manuel Le Gallo-bourdeau of Horgen (CH)

Kaoutar El Maghraoui of Yorktown Heights NY (US)

Malte Johannes Rasch of Chappaqua NY (US)

HsinYu Tsai of Cupertino CA (US)

CO-DESIGN OF A MODEL AND CHIP FOR DEEP LEARNING BACKGROUND - A simplified explanation of the abstract

This abstract first appeared for US patent application 20240289607 titled 'CO-DESIGN OF A MODEL AND CHIP FOR DEEP LEARNING BACKGROUND

Simplified Explanation:

This patent application describes a method, computer program product, and system for generating a processor design using a deep neural network. The processor selects an architecture search space and hardware components space, then optimizes the chip design and neural network for machine learning tasks.

• The processor selects an initial deep neural network and current chip design.
• It modifies the chip design using components from the hardware components space.
• The processor optimizes the neural network by selecting from the architecture search space.
• The optimized chip design and neural network are provided for machine learning tasks.

Key Features and Innovation:

• Generation of a processor design using a deep neural network.
• Selection of architecture search space and hardware components space.
• Optimization of chip design and neural network for machine learning tasks.

Potential Applications:

This technology can be applied in various fields such as artificial intelligence, machine learning, and computer hardware design.

Problems Solved:

This technology addresses the need for efficient processor design for executing deep neural networks.

Benefits:

• Improved performance in executing deep neural networks.
• Enhanced efficiency in processor design.
• Streamlined machine learning tasks.

Commercial Applications:

Potential commercial uses include AI hardware development, machine learning applications, and optimization of processor designs for specific tasks.

Prior Art:

Researchers can explore prior art related to deep neural network-based processor design and optimization methods in the field of computer hardware and machine learning.

Frequently Updated Research:

Researchers can stay updated on advancements in deep neural network-based processor design and optimization methods for machine learning tasks.

Questions about Deep Neural Network-based Processor Design:

1. What are the key components of a deep neural network-based processor design? 2. How does the optimization process improve the performance of the processor design?

Original Abstract Submitted

a method, computer program product, and system to generate a processor design via a deep neural network is provided. a processor selects an architecture search space and a hardware components space. a processor selects an initial deep neural network from the architecture search space. a processor determines an initial current chip design for executing the current deep neural network, wherein the initial chip design has a hardware performance metric for implementing the current deep neural network. a processor repeatedly executes an optimization method, the optimization method comprising modifying the chip design one or more times using components from the hardware components space and optimizing the current deep neural network by selecting a deep neural network from the architecture search space. a processor provides the optimized chip design and the specific deep neural network for performing the machine learning task.