Graphics Pipeline Time Multiplexing for Secure Cockpit Display
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current graphics processing devices lack the necessary security mechanisms to ensure secure processing and separation of graphic data from different applications, particularly in high-security environments like aircraft cockpits, where applications with varying security levels need to access shared resources without compromising security.
Innovation Solution
A graphics processing device with a GPU architecture, implemented as an FPGA, incorporates a graphics pipeline that allows for temporal separation of graphic objects from different applications by interrupting the processing of one application to process another, using time multiplexing and context data management to ensure secure and efficient processing, with modules like geometry and raster modules handling graphics data and pixel memory for secure spatial segregation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple applications with different security levels access the same graphics processing device simultaneously, then processing efficiency is improved, but security is compromised
Solution Approach 1:
The graphics processing device is segmented into multiple isolated graphics pipelines, where each pipeline is dedicated to processing graphic objects from a specific application. This spatial segmentation ensures that applications with different security levels are physically isolated, preventing security compromises while allowing simultaneous processing of multiple applications, thus maintaining high processing efficiency.
Solution Approach 2:
A control device acts as an intermediary between applications and graphics pipelines. It manages the acceptance of graphic data from applications, determines the security level of each application, and routes graphic objects to appropriate pipelines. This intermediary mechanism enables secure multi-application processing by enforcing security policies without reducing overall processing throughput.
2Adaptability or versatility
If a single graphics pipeline processes graphic objects from multiple applications, then device utilization is improved, but security separation is lost
Solution Approach 1:
The graphics processing device is divided into multiple independent graphics pipelines, each dedicated to a specific application. This segmentation maintains security separation while allowing each pipeline to be fully utilized by its assigned application, achieving both security isolation and high device utilization through parallel processing.
Solution Approach 2:
Each graphics pipeline is designed to be universal in its capability to process graphic objects, but specialized in its assignment to a particular application. The control device dynamically assigns applications to pipelines based on security levels, allowing the system to adapt to different utilization scenarios while maintaining security boundaries.
3Reliability
If graphic processing is interrupted and resumed later, then security context switching is achieved, but processing time increases
Solution Approach 1:
The control device preliminarily determines the security level of each application and pre-assigns graphic objects to appropriate graphics pipelines before processing begins. This preliminary action eliminates the need for context switching during processing, as each pipeline continuously processes graphic objects from its assigned application without interruption, thereby maintaining security separation while avoiding time losses.
Solution Approach 2:
Each graphics pipeline maintains continuous processing of graphic objects from its assigned application without interruption or context switching. This continuity of useful action ensures that security boundaries are maintained throughout the processing sequence while minimizing total processing time, as no time is lost to context switching operations.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a graphics processing device 1, comprising a control unit 2 for receiving graphical data related to first graphic objects from a first application App_1, and for receiving graphical data related to second graphic objects from a second application App_2; further comprising a geometry module 5 for generating the geometry data of the graphic objects from the graphical data; a raster module 7 for rasterizing the graphic objects based on the geometry data, and for generating pixels of the graphic objects, wherein the control unit 2, the geometry module 5, and the raster module 7 form a graphics pipeline configured to process one of the graphic objects in one serial pass through, during which pixels of the graphic object are generated from the graphical data for the graphic object. The graphic pipeline is configured to interrupt the processing of one of the first graphic objects of the first application App_1 in a first serial pass through so that a partially processed graphic object is obtained. At least one of the second graphic objects of the second application App_2 is processed in a second serial pass through, and subsequently, the processing of the partially processed graphic object of the first application App_1 is continued.