Split Anti-Aliased Sample Storage for Memory Bandwidth Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Anti-aliasing techniques, such as supersampling and multisampling, increase the memory footprint and memory bandwidth congestion due to the storage and retrieval of multiple samples per pixel, leading to performance degradations in image rendering.
Innovation Solution
Storing anti-aliased samples in different memories based on their likelihood of use, with frequently used samples in faster graphics memory and less frequently used samples in slower system memory, to optimize memory access and reduce memory traffic.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple anti-aliased samples are stored per pixel in memory, then image rendering quality is improved, but memory footprint increases substantially
Solution Approach 1:
The patent divides the set of anti-aliased samples into different categories or groups, storing them in different memory locations or memory types. This segmentation allows the system to manage memory resources more efficiently by not storing all samples uniformly, thus reducing overall memory footprint while preserving necessary rendering quality.
Solution Approach 2:
The patent applies different storage strategies to different samples based on their specific characteristics or usage patterns. Frequently accessed or more important samples are stored in faster memory, while less critical samples are stored in slower or less expensive memory, optimizing both quality and memory usage locally for each sample type.
2Measurement precision
If multiple samples are stored and accessed in memory for anti-aliasing, then rendering accuracy is improved, but memory bandwidth congestion increases
Solution Approach 1:
By segmenting samples into different storage locations, the patent enables parallel or selective access to different sample sets, reducing contention for memory bandwidth and improving overall memory utilization efficiency.
Solution Approach 2:
The patent introduces an intermediary mechanism (such as a sample selection unit or memory management structure) that manages access to multiple samples, coordinating memory operations to reduce bandwidth congestion and improve access efficiency.
3Speed
If all anti-aliased samples are stored in local graphics memory, then rendering speed is improved, but memory capacity is exceeded
Solution Approach 1:
The patent segments the sample storage across different memory hierarchies (local graphics memory and system memory), allowing the system to maintain high-speed access for critical samples while offloading less critical samples to system memory, thus avoiding local memory capacity overflow.
Solution Approach 2:
The patent extends the storage solution from a single memory dimension to multiple memory dimensions or levels, utilizing both local graphics memory and system memory as different layers in a memory hierarchy, thereby solving the capacity limitation while maintaining speed advantages.
Data Source
AI summary
Embodiments of the present invention are directed to improving the performance of anti-aliased image rendering. One embodiment is a method of rendering a pixel from an anti-aliased image. The method includes: storing a first set and a second set of samples from a plurality of anti-aliased samples of the pixel respectively in a first memory and a second memory; and rendering a determined number of said samples from one of only the first set or the first and second sets. Corresponding system and computer program product embodiments are also disclosed.