CPU-GPU Feature Information Batching for Visualization Frame Rates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for visualizing associated data on a GPU experience reduced frame rates due to frequent communication between the CPU and GPU, leading to impaired display performance, as each piece of feature information requires separate transmission.
Innovation Solution
Combining multiple pieces of feature information into overall feature information, which is then transmitted once from the CPU to the GPU, reducing the number of communication instances and enhancing GPU frame rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple pieces of feature information are transmitted separately from CPU to GPU, then each piece of data can be processed individually, but the communication frequency between CPU and GPU increases, reducing GPU frame rate
Solution Approach 1:
The patent combines multiple separate feature information transmissions into a single batched transmission. Instead of sending each feature information piece individually from CPU to GPU, the system accumulates multiple pieces and transmits them together in one operation, thereby reducing communication frequency and improving GPU frame rate while maintaining complete data processing capability
2Loss of time
If communication between CPU and GPU occurs frequently, then data can be processed in real-time, but the frame rate is reduced due to hardware limitations
Solution Approach 1:
The patent implements periodic batched communication between CPU and GPU instead of continuous individual transmissions. The system processes multiple feature information pieces during CPU execution time and performs GPU data processing periodically in batches, reducing communication overhead frequency while maintaining efficient real-time processing capability
Data Source
AI summary
Disclosed herein are methods and systems for improving GPU processing during visualization operations. In one embodiment, the method comprises receiving a data frame at a central processing unit (CPU), the data frame including a plurality of pieces of associated data; identifying, by the CPU, feature information corresponding to the plurality of pieces of associated data; generating, by the CPU, overall feature information by combining at least a portion of the feature information; and sending, by the CPU, the overall feature information to a graphics processing unit (GPU).