Graphics Processor Unit Periodic Reset for Safety-Critical Rendering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing graphics processing units (GPUs) in safety-critical systems face challenges in meeting stringent safety standards due to random errors caused by transient events like ionizing radiation, voltage spikes, and electromagnetic pulses, which can lead to incorrect rendering of critical information, and existing dual lockstep processors are costly and power-consuming.
Innovation Solution
Implementing periodic resets of the GPU with a safety controller that schedules resets based on a predefined frequency to mitigate transient errors, including soft and hard resets to reinitialize processing elements and invalidate cache entries, while monitoring safety metrics to adapt the reset frequency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dual lockstep processors are used to detect and correct transient errors, then reliability is improved, but device complexity and power consumption increase
Solution Approach 1:
The patent applies preliminary action by performing periodic resets of the graphics processing unit before transient errors can persist and cause safety violations. The safety controller schedules and executes resets at predetermined intervals, proactively preventing error accumulation rather than reacting to detected failures. This approach ensures reliability through preventive maintenance rather than through complex redundant hardware configurations.
Solution Approach 2:
The patent changes the operational parameters of the graphics processing unit by periodically resetting its internal state, cache, and buffers. This parameter change clears any transient errors that may have occurred during operation, allowing the system to maintain reliability without requiring dual lockstep processor configurations. The reset frequency and scope are adjusted based on safety requirements.
2Reliability
If dual lockstep processors are used to ensure safety standards, then reliability is improved, but power consumption increases
Solution Approach 1:
The safety controller performs preliminary resets of the graphics processing unit at scheduled intervals to prevent transient errors from causing safety violations. This proactive approach ensures compliance with safety standards like ISO 26262 without requiring continuous operation of power-consuming redundant hardware. The periodic resets clear potential errors before they can affect system safety.
Solution Approach 2:
The patent extracts the safety assurance function from the graphics processing unit itself and places it in a separate safety controller. The safety controller monitors the GPU and performs resets independently, allowing the GPU to operate at full performance without the overhead of integrated dual lockstep configurations. This separation enables safety compliance with reduced power consumption.
3Reliability
If periodic resets are performed frequently to clear transient errors, then reliability is improved, but productivity decreases
Solution Approach 1:
The safety controller implements periodic resets of the graphics processing unit at optimized intervals that balance reliability and productivity. Rather than continuous or overly frequent resets, the system performs resets at predetermined periods that are sufficient to clear transient errors but infrequent enough to maintain high graphics processing throughput. This periodic action ensures safety while minimizing impact on performance.
Solution Approach 2:
The reset frequency and scope are made dynamic rather than static. The safety controller can adjust the periodicity and extent of resets based on system state, error rates, and performance requirements. This dynamic approach allows the system to increase reset frequency when reliability concerns arise while maintaining higher productivity during normal operation, optimizing the trade-off between the two parameters.
Data Source
AI summary
A method of performing safety-critical rendering at a graphics processing unit within a graphics processing system, the method comprising: receiving, at the graphics processing system, graphical data for safety-critical rendering at the graphics processing unit; scheduling at a safety controller, in accordance with a reset frequency, a plurality of resets of the graphics processing unit; rendering the graphical data at the graphics processing unit; and the safety controller causing the plurality of resets of the graphics processing unit to be performed commensurate with the reset frequency.