DMA Internal Buffering for Scatter-Gather Line Memory Access
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing processing systems for self-driving and automated driving applications face inefficiencies in handling and manipulating arrays of data from multiple sensors, particularly in managing unaddressable portions of memory during scatter and gather operations.
Innovation Solution
A system utilizing a direct memory access (DMA) controller that loads addressable lines of data to a buffer, updates unaddressable portions, and initiates operations to write modified data to memory, independent of a memory controller, enhancing data access and processing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a memory controller is used to manage data access between DMA controller and memory, then memory access control is achieved, but processing latency increases and resource consumption increases
Solution Approach 1:
The patent extracts the buffer management functionality from the memory controller and relocates it to the DMA controller. The DMA controller now independently loads addressable lines from memory to its internal buffer, updates unaddressable portions, and writes modified data back to memory without requiring memory controller intervention, thereby reducing processing latency while maintaining reliable memory access control
Solution Approach 2:
The DMA controller is enhanced to perform self-service buffer management operations. It autonomously loads data lines from memory, modifies unaddressable portions within its internal buffer, and writes back to memory independently, eliminating the need for external memory controller management and reducing overall system resource consumption
2Reliability
If the DMA controller accesses memory through the bus for each operation, then memory access is achieved, but processing speed decreases and resource consumption increases
Solution Approach 1:
The DMA controller performs preliminary actions by loading entire addressable lines from memory to its internal buffer in advance. This allows subsequent updates to unaddressable portions to be performed locally in the buffer without repeated bus accesses, significantly improving processing speed while maintaining reliable memory access through the initial bulk load operation
3Reliability
If unaddressable portions of memory are updated through traditional methods, then data modification is achieved, but resource consumption increases and processing efficiency decreases
Solution Approach 1:
The patent introduces the DMA controller's internal buffer as an intermediary between memory and the processing system. The buffer serves as a local workspace where unaddressable portions can be updated without direct memory controller involvement or repeated bus transactions, achieving reliable data modification while reducing system resource consumption and energy loss
Data Source
AI summary
Example methods, apparatuses, and/or articles of manufacture are disclosed that may be implemented, in whole or in part, in connection with direct memory access (DMA) techniques. In one application, DMA techniques may be augmented by a buffer to enable improved servicing of scatter and gather requests.