Multicore Video Processing Allocation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video processing systems with multiple cores inefficiently allocate video unit data, leading to processing delays and increased power consumption due to the lack of consideration for core performance and video complexity, resulting in suboptimal decoding times and reduced efficiency in parallel processing.
Innovation Solution
A method for efficiently processing videos in parallel by obtaining parallelized unit information and performance information of multiple cores, and allocating video unit data based on complexity and performance, allowing for optimized tile allocation across asymmetric multicores, thereby improving decoding speed and reducing power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If video unit data is allocated without considering core performance and video complexity, then allocation simplicity is maintained, but processing speed decreases and power consumption increases
Solution Approach 1:
The system performs preliminary analysis of video complexity and core performance before allocation, obtaining parallelized unit information and performance information in advance. This allows the allocator to make informed decisions about assigning video units to cores based on pre-computed complexity metrics and core capabilities, resolving the contradiction by preparing allocation data beforehand rather than making simple arbitrary assignments.
Solution Approach 2:
The system changes the allocation parameters by incorporating video complexity metrics and core performance characteristics into the allocation decision process. Instead of using a fixed or random allocation scheme, the system dynamically adjusts allocations based on measured parameters such as parallelized unit information, complexity information, and core performance data, thereby improving processing speed while managing complexity through parameterized control.
2Use of energy by moving object
If video unit data is allocated without considering core performance and video complexity, then system complexity is reduced, but power consumption increases
Solution Approach 1:
The system monitors and responds to power consumption by dynamically adjusting allocation parameters. When power consumption exceeds thresholds or efficiency metrics deteriorate, the system modifies allocation decisions based on updated complexity and performance information, enabling energy-aware scheduling that balances power usage against processing requirements through parameterized control mechanisms.
Solution Approach 2:
The system implements feedback loops that monitor power consumption, processing progress, and core performance metrics during video decoding. This feedback information is used to adjust subsequent allocation decisions, allowing the system to optimize power consumption by redistributing workloads based on real-time performance data and power efficiency measurements, thereby resolving the contradiction between energy efficiency and allocation complexity.
3Productivity
If complex allocation strategies are used to optimize processing, then processing speed improves, but system complexity increases
Solution Approach 1:
The system segments the video decoding task into independent parallelizable units and separately segments the allocation decision-making into distinct phases: obtaining parallelized unit information, obtaining performance information, and making allocation decisions. This segmentation allows complex optimization to be applied to individual components without overwhelming system complexity, as each segment can be processed and optimized independently.
Solution Approach 2:
The system introduces intermediary data structures and metrics (parallelized unit information, complexity information, performance information) that mediate between the raw video data and core allocation decisions. These intermediaries simplify the allocation process by pre-computing and structuring the necessary information, allowing complex optimization logic to operate on well-defined parameters rather than raw data, thereby managing system complexity while maintaining high decoding efficiency.
Data Source
AI summary
According to one embodiment of the present disclosure, there is provided a processing method including obtaining parallelized unit information for video parallel processing; obtaining performance information that indicates performance of a plurality of cores; and allocating a plurality of video unit data to the plurality of cores based on the parallelized unit information and the performance information. The parallelized unit information includes complexity information that indicates complexity of the plurality of video unit data.