Mixed Core Processor Unit with Neural and Digital Cores
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data processing systems face challenges in efficiently processing large amounts of data due to slower processing speeds and higher energy consumption, particularly in devices like automobiles and satellites where power constraints are limiting.
Innovation Solution
A processor unit comprising a group of neural cores and digital processing cores connected by a routing network, allowing for both analog and digital data processing, which increases precision and processing power while optimizing energy use.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If supercomputers are used to process large amounts of data, then processing speed is improved, but energy consumption increases
Solution Approach 1:
The processor is divided into multiple specialized cores (neural cores for analog processing, digital processing cores for digital operations) that can handle different types of computations simultaneously. This segmentation allows the system to process large amounts of data efficiently without requiring the full power of a supercomputer, thereby reducing energy consumption while maintaining high productivity.
Solution Approach 2:
Different types of cores are assigned to different processing tasks based on their specialized capabilities. Neural cores with analog circuitry handle neural network computations locally, while digital processing cores handle other digital operations. This local specialization optimizes energy usage for each specific processing task rather than using a uniform high-power approach for all operations.
2Productivity
If supercomputers are used to process large amounts of data, then processing capability is improved, but device complexity increases
Solution Approach 1:
The patent combines multiple types of processing cores (neural and digital) into a single integrated processor unit. This merging allows the system to achieve supercomputer-level processing capability for specific tasks while avoiding the complexity of multiple separate systems, as all cores are managed within one unified processor architecture.
Solution Approach 2:
The processor is designed with multi-functional cores that can handle various types of computations. Neural cores can process different neural network operations, while digital processing cores can execute various digital instructions. This universality allows a single processor to perform multiple functions that would otherwise require separate specialized systems, reducing overall device complexity.
3Use of energy by moving object
If analog processing is used in neural cores, then energy efficiency is improved, but measurement precision may be reduced
Solution Approach 1:
Analog-to-digital converter circuits serve as intermediaries between the analog neural cores and the digital processing cores. These converters translate analog signals from the energy-efficient neural processing into digital format, preserving precision while maintaining the energy advantages of analog processing. This intermediary approach allows the system to benefit from both analog energy efficiency and digital precision.
Data Source
AI summary
A method and apparatus for processing data. The data is sent to a processor unit comprising a group of neural cores, a group of digital processing cores, and a routing network connecting the group of digital processing cores. The data is processed in the processor unit to generate a result.