Compressed Memory Zones With Variable Slots to Prevent Fragmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory compression techniques incur additional computational costs and performance degradation due to decompression overhead and may cause memory fragmentation.
Innovation Solution
A memory device with compressed zones and a controller that selectively compresses data based on compressibility, storing it in slots of varying sizes to minimize overhead and prevent fragmentation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data is compressed using traditional techniques (LZ77, Huffman coding, RLE), then memory capacity is increased, but computational overhead and decompression time increase
Solution Approach 1:
The memory array is divided into multiple compressed zones, each with slots of different sizes. This segmentation allows data elements of varying compressed sizes to be stored in appropriately-sized slots, improving space utilization while enabling parallel access to multiple zones, thereby reducing decompression time overhead.
Solution Approach 2:
Different zones are allocated with different slot sizes based on local requirements. Frequently accessed or highly compressible data can be placed in zones with smaller slot sizes for faster access, while less frequently accessed data uses larger slots. This local optimization balances compression efficiency with access speed.
2Device complexity
If data is compressed into fixed-size slots, then memory structure is simplified, but memory fragmentation occurs
Solution Approach 1:
The patent employs slots of asymmetric (varying) sizes within different compressed zones rather than uniform fixed-size slots. This asymmetric structure allows compressed data elements of different sizes to fit more naturally into appropriate slots, reducing wasted space and preventing memory fragmentation while maintaining manageable structural complexity through organized zone divisions.
3Productivity
If compression is applied to all data, then memory bandwidth is increased, but computational cost increases
Solution Approach 1:
The system changes the compression parameter by using variable slot sizes across different zones rather than applying uniform compression. This allows the memory system to adapt compression levels and slot allocations based on data characteristics, achieving high memory bandwidth for compressible data while avoiding unnecessary computational costs for data that doesn't benefit from compression.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A memory device includes a memory array including compressed zones including a first compressed zone including slots of a first slot size and a second compressed zone including slots of a second slot size; and a controller configured to compress a first data element received from a processing block and store the compressed first data element in a first slot of the slots of the first compressed zone based on a data size of the compressed first data element.