General Purpose Stored Program Computer TRIZ Cases
- CNN Accelerator Internal Buffering for Pooling Data Reuse
Convolutional neural network accelerator uses internal buffers to store intermediate pooling results within processing elements.
- Preboot Language Agent for Centralized Imaging Management
A preboot language agent interprets specification files to execute generalized imaging tasks across diverse client devices.
- Network-on-Chip Switch Routing via Direction Determinator
A network switch determines routing direction using location and destination coordinates to select output links.
- Reconfigurable Memory Module for Stream Processing Pipeline Balance
Dynamic memory sizing in a reconfigurable module resolves pipeline imbalance and reduces external bandwidth consumption during graphic processing.
- Electronic Location Code for Compact Data Transmission
Segmented location codes balance measurement precision against data volume by dynamically adjusting bit allocation across unified sub-segments.
- Website Component Parameterized Behavior Override
Element handler processes override requests between website components to manage complex interactions without increasing system structure complexity.
- Sticky Bit Segmentation for Speculative Execution
Segmented sticky bits resolve register renaming contradictions by reducing storage requirements while maintaining state restoration capability.
- Howard Cascade Inter-Chip Communication via Shared Memory Queues
Howard Cascade architecture routes data through input queues and switch fabric, eliminating memory locking delays and blocking in multiprocessor systems.