Apple inc. (20240330217). INPUT AND OUTPUT SPATIAL CROPPING OPERATIONS IN NEURAL PROCESSOR CIRCUITS simplified abstract
Contents
INPUT AND OUTPUT SPATIAL CROPPING OPERATIONS IN NEURAL PROCESSOR CIRCUITS
Organization Name
Inventor(s)
Christopher L. Mills of Cupertino CA (US)
INPUT AND OUTPUT SPATIAL CROPPING OPERATIONS IN NEURAL PROCESSOR CIRCUITS - A simplified explanation of the abstract
This abstract first appeared for US patent application 20240330217 titled 'INPUT AND OUTPUT SPATIAL CROPPING OPERATIONS IN NEURAL PROCESSOR CIRCUITS
Simplified Explanation: The patent application describes a social circuit that includes a neural processor circuit connected to a CPU. The neural processor circuit consists of neural engines, a data processor DMA circuit, a system memory, and a data processor circuit. The CPU runs a compiler that determines the mode of spatial cropping and the crop offset. The neural processor circuit supports arbitrary cropping in the x and y dimensions. The compiler generates task descriptors distributed to the components of the neural processor circuit. The data processor DMA circuit fetches and formats data corresponding to the crop from a source to the buffer. The buffer realigns the data according to the crop origin for broadcast to the neural engines, which perform computation operations using the cropped data.
- Neural processor circuit with neural engines
- CPU executing a compiler for spatial cropping
- Arbitrary cropping in x and y dimensions
- Data processor DMA circuit for fetching and formatting data
- Buffer realigning data for broadcast to neural engines
Potential Applications: - Image and video processing - Artificial intelligence and machine learning tasks - Robotics and automation systems
Problems Solved: - Efficient data processing for neural networks - Improved spatial cropping capabilities - Enhanced performance in computation operations
Benefits: - Faster and more accurate data processing - Increased flexibility in cropping dimensions - Optimal utilization of neural processor circuit components
Commercial Applications: Title: Advanced Neural Processor Circuit for Enhanced Data Processing This technology can be applied in industries such as: - Healthcare for medical imaging analysis - Automotive for autonomous driving systems - Security for surveillance and facial recognition systems
Questions about Neural Processor Circuit Technology: 1. How does the neural processor circuit improve data processing efficiency? 2. What are the key advantages of using arbitrary cropping in the x and y dimensions for neural networks?
Original Abstract Submitted
an soc circuit includes a neural processor circuit coupled to a cpu. the neural processor circuit includes neural engines, a data processor dma circuit, a system memory, and a data processor circuit. the cpu is configured to execute a compiler, which is in turn configured to determine to perform a mode of spatial cropping and the associated crop offset. the neural processor circuit is configured to support arbitrary cropping in the x and y dimensions. the compiler is configured to generate task descriptor(s), the task descriptor(s) distributed to components of the neural processor circuit. the data processor dma circuit is configured to fetch and format data corresponding to the crop from a source to the buffer. the buffer is configured to realign the data according to the crop origin for broadcast to the neural engines. the neural engines is configured to perform a computation operation which uses the cropped data.
