Dynamic CPU GPU Graphics Processing Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current graphics processing technologies cannot accurately determine whether a central processing unit (CPU) or a graphics processing unit (GPU) will perform graphics processing faster, as the time required for graphics processing by the CPU varies with the number of screen elements and pixel count, while the GPU's time remains relatively constant, leading to inefficient resource allocation.
Innovation Solution
A graphics processing device and method that includes a CPU and a GPU, where the CPU judges which device to use for graphics processing based on the number of screen elements, performing the processing itself if faster and delegating to the GPU after draw call batching if the GPU is faster.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If the CPU performs graphics processing, then the processing time is shorter when the number of screen elements is low, but the processing time increases in proportion to the pixel count of the display
Solution Approach 1:
The patent applies dynamics by making the graphics processing device dynamically switchable between CPU and GPU based on real-time conditions. The system determines the appropriate processing device by evaluating both the number of screen elements and pixel count, allowing adaptive selection rather than a fixed architecture. This dynamic approach resolves the contradiction by enabling the system to optimize for speed (CPU for low element counts) while maintaining adaptability to different display resolutions (GPU for high pixel counts).
2Loss of time
If the GPU performs graphics processing, then the processing time is almost constant regardless of pixel count, but additional draw call batching time is required
Solution Approach 1:
The patent applies preliminary action by performing draw call batching as a preparatory step before GPU processing. The system combines attributes of all screen elements (textures, materials, coordinates) into a single batch of data in advance, which the GPU then processes efficiently. This preliminary organization of data resolves the contradiction by accepting the additional batching step as a trade-off for the GPU's constant-time processing capability across different pixel counts.
3Productivity
If the processing device is selected based only on pixel count, then high-resolution displays can utilize GPU advantages, but the actual processing speed cannot be optimized because screen element count also significantly affects CPU processing time
Solution Approach 1:
The patent applies parameter changes by expanding the selection criteria from a single parameter (pixel count) to multiple parameters (screen element count and pixel count). The system evaluates both parameters simultaneously to determine whether CPU or GPU processing will be faster. This multi-parameter approach resolves the contradiction by enabling accurate productivity optimization through precise measurement of actual processing time requirements based on the specific combination of screen elements and display resolution.
Data Source
AI summary
A graphics processing device includes: a CPU corresponding to a general-purpose; and a GPU corresponding to a special-purpose for graphics processing, the GPU being configured to necessitate draw call batching before the GPU starts its operation, wherein: the CPU judges which device, the CPU itself or the GPU, should perform graphics processing to generate a screen image to be displayed on a display with reference to the number of screen elements composing the screen image; and the CPU performs graphics processing to generate the screen image if the CPU judges that the CPU itself should perform graphics processing, and the CPU performs draw call batching then makes the GPU perform graphics processing to generate the screen image if the CPU judges that the GPU should perform graphics processing.