Video Workload Allocation via Hardware Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video streaming systems face challenges in determining an ideal hardware configuration for video processing workloads, leading to either insufficient resources causing poor user experience or idle resources due to over-provisioning.
Innovation Solution
A point system is used to pair hardware configurations with video processing workloads, assigning point values to both hardware systems and workloads based on their capabilities and requirements, allowing for dynamic assignment of workloads to suitable systems, including the option of using a neural network to refine these assignments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If large computers with sufficient hardware resources are provided, then video processing performance is improved and real-time encoding/decoding is enabled, but hardware resources become idle due to over-provisioning
Solution Approach 1:
The patent implements dynamic workload management by continuously monitoring system utilization metrics and adjusting workload distribution in real-time. The load manager dynamically assigns video processing workloads to appropriate systems based on current resource availability, preventing both over-provisioning and under-provisioning. This dynamic adjustment allows the system to optimize performance while minimizing idle hardware resources.
Solution Approach 2:
The system changes operational parameters by using utilization thresholds to control workload assignment. When system utilization falls below a threshold, new workloads are assigned to that system; when it exceeds the threshold, workloads are redirected elsewhere. This parameter-based control enables efficient resource utilization while maintaining video processing performance.
2Loss of energy
If hardware resources are reduced to avoid idle resources, then resource efficiency is improved, but video streams may skip or stutter causing poor user experience
Solution Approach 1:
The patent implements a feedback mechanism where the load manager continuously monitors system utilization metrics and uses this feedback to make intelligent workload assignment decisions. By tracking utilization thresholds and adjusting assignments based on real-time conditions, the system ensures that video processing workloads are always assigned to systems with sufficient capacity to maintain stream quality, preventing skips or stuttering while optimizing resource efficiency.
Solution Approach 2:
The system performs preliminary assessment of hardware capabilities and workload requirements before assignment. The load manager evaluates system utilization thresholds and workload characteristics in advance, ensuring that workloads are only assigned to systems that can reliably handle them. This preliminary action prevents poor user experience by avoiding assignment of complex workloads to under-provisioned systems.
3Device complexity
If manual configuration of hardware settings is used, then system complexity is reduced, but adaptability to different video processing workloads is limited
Solution Approach 1:
The patent implements self-service through automated workload assignment. The load manager automatically evaluates system utilization metrics and assigns video processing workloads to appropriate systems without manual intervention. The system monitors its own state and makes intelligent decisions about workload distribution, providing both simplicity for users and high adaptability to different workload conditions.
Solution Approach 2:
The load manager serves multiple functions: it monitors system utilization, evaluates workload requirements, makes assignment decisions, and adjusts to changing conditions. This universal component handles diverse video processing workloads across multiple systems with varying capabilities, providing both ease of operation and adaptability through a single multi-functional system.
4Measurement precision
If complex rules engine or neural network is used to assign workloads, then workload assignment accuracy is improved, but system complexity increases
Solution Approach 1:
The patent segments the workload assignment problem into distinct components: hardware scoring, workload scoring, and threshold-based matching. By dividing the complex decision-making process into separate scoring functions and comparison logic, the system achieves high assignment accuracy while maintaining manageable complexity. Each component can be independently optimized and understood.
Data Source
AI summary
A particular method includes determining, at a processor of a computing device, a plurality of hardware scores associated with a plurality of hardware systems, where each of the plurality of hardware scores corresponds to a respective one of the plurality of hardware systems. The method also includes receiving data indicating a video processing workload and determining a workload score associated with the video processing workload based on the data. The method further includes identifying one or more of the plurality of hardware systems to execute the video processing workload. The one or more hardware systems are identified based on a comparison of the workload score to one or more hardware scores of the one or more hardware systems. The method includes assigning the one or more hardware systems to execute the video processing workload.