Hybrid Scoreboard for Power-Efficient Video Thread Dependency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Multi-thread decoding of video information consumes significant power due to software scoreboard-based approaches, which are inefficient in managing thread dependencies and power usage.
Innovation Solution
A hybrid scoreboard-based approach combining hardware and software scoreboards to enforce thread dependencies, where threads with long waiting times are processed through hardware scoreboards and those with short waiting times through software scoreboards, minimizing software polling and optimizing power efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If software polling is utilized to verify completed/uncompleted tasks in a software scoreboard-based approach, then thread dependency management is achieved, but power consumption increases significantly
Solution Approach 1:
The patent segments the thread dependency management into two distinct systems: a hardware scoreboard for tracking completed tasks and a software scoreboard for managing thread dependencies. This segmentation allows each system to specialize in its function, with the hardware scoreboard handling power-intensive polling tasks and the software scoreboard handling dependency logic, thereby reducing overall power consumption while maintaining reliability.
Solution Approach 2:
The patent introduces a hybrid scoreboard mechanism that acts as an intermediary between hardware and software systems. The hardware scoreboard provides status information to the software scoreboard, which then manages thread dependencies. This intermediary approach allows the system to leverage hardware speed for status tracking while using software flexibility for dependency management, reducing the need for continuous software polling and thus lowering power consumption.
2Use of energy by moving object
If hardware scoreboard is used for thread dependency enforcement, then power consumption is reduced, but the number of handleable dependencies is limited
Solution Approach 1:
The patent divides the dependency management workload between hardware and software scoreboards. The hardware scoreboard handles a subset of dependencies with fixed entries, providing low-power operation for these critical dependencies. The software scoreboard handles the remaining dependencies, providing flexibility and adaptability. This segmentation allows the system to achieve both low power consumption and high adaptability by distributing the workload appropriately.
Solution Approach 2:
The hybrid scoreboard system provides multi-functionality by combining the fixed-entry hardware scoreboard with the flexible software scoreboard. The hardware scoreboard handles dependencies requiring deterministic timing and low power consumption, while the software scoreboard handles dependencies requiring flexibility and adaptability. Together, they form a universal system that can handle a large number of dependencies with varying requirements, achieving both power efficiency and versatility.
3Reliability
If software polling is performed frequently to check thread completion, then thread execution accuracy is maintained, but processing time and power usage increase
Solution Approach 1:
The hardware scoreboard performs preliminary action by continuously tracking and recording the completion status of threads in advance. This pre-computed status information is made available to the software scoreboard without requiring frequent polling. The software scoreboard can then query the hardware scoreboard for completion status, reducing the frequency and duration of polling operations, thereby decreasing processing time while maintaining execution accuracy.
Solution Approach 2:
The system implements feedback mechanisms where the hardware scoreboard provides status feedback to the software scoreboard about thread completion. This feedback loop allows the software scoreboard to update its knowledge of thread status without needing to perform exhaustive polling. The feedback mechanism reduces the time required to determine thread completion status while maintaining accurate tracking, thus reducing processing time without sacrificing reliability.
Data Source
AI summary
Described herein are technologies related to enforcing thread dependency using a hybrid scoreboard. An encoded video information that includes a plurality of threads is received, a first set and a second set of threads from the plurality of thread is determined, the first and second sets of threads are assigned to a hardware and a software, respectively, and dependency threads in the first and second sets of threads is enforced.