Shared-Buffer Video Transcoding for Low-Latency Encoding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video transcoding systems face challenges in minimizing delay and power consumption, particularly in battery-powered devices used for videoconferencing, due to the inefficiencies in the process of decoding and re-encoding video data, which involve significant latency and resource consumption.
Innovation Solution
Implementing a transcoding process where decoding and encoding operations are synchronized directly between accelerated processing devices without mediation by the host operating system, utilizing buffer looping to reduce the need for memory round-trips, thereby reducing latency and power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If video data is decoded and re-encoded through memory storage, then the transcoding process can be completed, but the latency increases and power consumption rises
Solution Approach 1:
The patent merges the decode and encode operations into a single unified process by sharing the buffer memory between the decode and encode accelerated processing devices. This eliminates the separate memory read and write operations, reducing both latency and power consumption while maintaining the transcoding functionality.
Solution Approach 2:
The shared buffer acts as an intermediary between the decode and encode operations, allowing direct data transfer without involving the host operating system or additional memory operations. This mediator enables simultaneous access by both devices, reducing the overall processing time and energy requirements.
2Productivity
If memory round-trips are used for decoding and encoding, then data can be processed, but the power consumption increases significantly
Solution Approach 1:
The patent combines the memory resources of both decode and encode operations into a single shared buffer, eliminating redundant memory operations. This maintains full transcoding throughput while significantly reducing the energy consumed by memory access operations.
3Loss of time
If host operating system mediation is used for decoding and encoding, then resource management is simplified, but latency increases
Solution Approach 1:
The patent extracts the buffer management functionality from the host operating system and implements it directly in the accelerated processing devices. This removes the OS mediation layer, reducing latency while the shared buffer protocol maintains manageable system complexity through direct device-to-device communication.
Data Source
AI summary
A technique for performing video transcoding is provided. A decoding device receives source video data in a first coding scheme. The decoding device writes one or more blocks of decoded video data to a buffer in response to the information provided by the encoding device. The encoding device encodes video data written to the buffer; wherein the encoding results in a block of encoded video data in a second coding scheme different from the first coding scheme. The writing and encoding steps are repeated for a plurality of further iterations, wherein each further iteration is initiated in response to encoding a block of video data.