Vehicle Multi-SoC Display Control With Dynamic Workload Sharing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing domain control systems in vehicles face challenges with insufficient hardware headroom for adding new functions and cannot efficiently manage high performance requirements, particularly when controlling a large number of displays, as they often rely on a single ECU which is insufficient for handling increased demands.
Innovation Solution
A vehicle system utilizing multiple System on Chips (SoCs) connected through a PCIe interface, where SoCs can request resource assistance from each other, terminate applications based on priority, and share workload to manage display operations, enabling efficient resource distribution and upgrade capabilities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a single ECU is used to control displays, then device complexity is reduced, but the system cannot meet high performance requirements when controlling a large number of displays
Solution Approach 1:
The system divides the display control function into multiple independent SoC units, where each SoC can independently control one or more displays. This segmentation allows the system to scale display control capability by adding more SoC units without requiring a single complex ECU, thus resolving the contradiction between productivity and device complexity.
Solution Approach 2:
Each SoC unit is designed with universal functionality to control displays, allowing any SoC to take over control tasks regardless of which display needs to be controlled. This multi-functionality enables flexible workload distribution across multiple SoCs, improving display control capability while maintaining relatively simple individual unit design.
2Adaptability or versatility
If hardware resources are increased to support new functions, then adaptability is improved, but hardware headroom becomes insufficient
Solution Approach 1:
The system segments hardware resources into multiple SoC units, each capable of independently executing applications and controlling displays. This segmentation allows new functions to be added by distributing workload across existing SoC units or adding new SoC units, rather than requiring a single large ECU with sufficient headroom, thus improving adaptability while managing hardware resources efficiently.
Solution Approach 2:
The system implements dynamic workload distribution where SoC units can request and receive applications from other SoCs based on their current resource state. This dynamic resource allocation allows the system to adapt to new functions and display requirements without requiring excessive static hardware headroom in each unit, resolving the contradiction between adaptability and hardware headroom.
3Productivity
If workload is concentrated in one SoC, then device complexity is reduced, but resource utilization becomes inefficient when multiple displays need control
Solution Approach 1:
The system implements a feedback mechanism where each SoC monitors its own resource state (busy/idle) and communicates this information to other SoCs. When an SoC is busy, it can request applications from other SoCs, and the requesting SoC receives feedback about resource availability. This feedback-driven workload distribution optimizes resource utilization across the system while maintaining manageable complexity through standardized communication protocols.
Data Source
AI summary
Proposed are a device provided to a vehicle and a control method therefor. Specifically, proposed is a device provided to a vehicle, the device comprising: a plurality of system on chips (SoCs) which execute at least one application and are connected via a predetermined input/output interface; and a display which is provided in the vehicle and outputs an execution screen of the at least one application. Further, proposed is that a first SoC, among the plurality of SoCs, requests the execution of a specific application, among the at least one application, to a second SoC on the basis of the resource state of the first SoC, and as a response to the request, receives the execution result of the specific application from the second SoC.