3D Flash Memory Tile Group Access and Data Restructuring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current memory access techniques in 3D flash memory devices require multiple accesses to different blocks, leading to increased access time and energy consumption due to the need to input block and tile addresses separately for each sector, reducing the usability and efficiency of memory devices.
Innovation Solution
Implementing independently selective tile group access methods that allow simultaneous access to multiple tile blocks using a data restructuring module to modify the order of data access, enabling faster and more energy-efficient memory operations by disabling unnecessary tiles and optimizing data transfer sequences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple blocks are accessed sequentially using present memory access schemes, then each block can be accessed individually, but access time increases and energy consumption increases due to requiring multiple separate accesses
Solution Approach 1:
The patent combines multiple tile block accesses into a single unified access operation. By merging the access requests for multiple blocks (e.g., Block 0 and Block 1) into one simultaneous access, the system eliminates the sequential overhead of separate accesses, thereby reducing access time while maintaining individual block accessibility.
Solution Approach 2:
The memory device is segmented into independently selectable tile groups within blocks, where each tile block can be independently addressed and accessed. This segmentation allows the system to selectively access only the necessary tile blocks simultaneously, improving access speed by avoiding unnecessary sequential operations while maintaining granular control over memory access.
2Productivity
If multiple blocks are accessed sequentially using present memory access schemes, then each block can be accessed individually, but energy consumption increases due to requiring multiple separate accesses
Solution Approach 1:
The patent merges multiple block access operations into a single unified access operation. By combining the access requests for multiple blocks into one simultaneous operation, the system reduces the total number of access cycles required, thereby lowering energy consumption associated with repeated access setup and execution.
Solution Approach 2:
The system dynamically selects and activates only the specific tile blocks that need to be accessed, rather than accessing all blocks sequentially. This dynamic selection mechanism allows the memory device to adapt to actual access patterns, activating necessary blocks simultaneously while keeping others inactive, thus optimizing energy efficiency.
3Quantity of substance
If block size is increased by stacking multiple layers to increase bit density, then storage capacity increases, but usability decreases due to reduced accessibility of individual sectors
Solution Approach 1:
The patent segments each block into multiple independently selectable tile groups, where each tile block contains specific memory elements. This segmentation allows the system to access individual sectors or subsets of sectors within a block without requiring access to the entire block, thereby maintaining high usability and ease of operation despite increased bit density from stacking multiple layers.
Data Source
AI summary
Embodiments of the present disclosure include data structuring techniques and configurations for memory access. In one embodiment, an apparatus includes a plurality of tiles, a plurality of blocks, wherein individual tiles of the plurality of tiles and individual blocks of the plurality of blocks each include a plurality of tile blocks having memory elements and wherein the plurality of tile blocks are accessible for read or write according to a tile address to identify a tile of the plurality of tiles and a block address to identify a block of the plurality of blocks and a data restructuring module configured to modify, based on the tile address, an order of data to be read from or written to the tile blocks. Other embodiments may be described and/or claimed.