Compressed Memory Writes to Cut NVRAM Rank Write Energy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing non-volatile random access memory (NVRAM) devices, such as FLASH memory and emerging technologies like memristor and phase change RAM, consume write energy when storing data, even though they do not require continuous power to maintain data persistence, and existing SRAM and DRAM require continuous power to retain data.
Innovation Solution
The techniques involve compressing data blocks before writing to a rank of memory devices, using a compression header and error correction code (ECC) to determine if data is compressed, and only writing to the necessary memory devices, thereby reducing write energy consumption, and also include encryption and decryption mechanisms to enhance security and energy efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If data is written to all memory devices in a rank, then data storage capacity is maximized, but write energy consumption increases
Solution Approach 1:
The patent divides the memory rank into multiple memory devices and further segments data into data units that can be selectively written to specific memory devices. By segmenting the write operation to only affected memory devices rather than writing to all devices in the rank, the system reduces write energy consumption while maintaining the ability to store the full data set across the memory rank.
2Use of energy by moving object
If compression is applied to data blocks, then write energy is reduced, but device complexity increases due to compression headers and ECC requirements
Solution Approach 1:
The patent applies compression to data blocks before writing them to memory devices. By performing the compression action in advance during the write operation, the system reduces the amount of data that needs to be written, thereby reducing write energy consumption. The compression header is added as metadata to track the compressed state, enabling efficient read operations without requiring decompression of entire data blocks.
3Reliability
If encryption is implemented for data security, then security is enhanced, but write energy consumption increases
Solution Approach 1:
The patent applies encryption selectively to only those data units that require security protection, rather than encrypting all data uniformly. By identifying and encrypting only the necessary portions of data locally, the system enhances security for sensitive information while minimizing the additional write energy consumption that would result from encrypting entire data blocks or all memory devices.
Data Source
AI summary
Techniques for writing data to a subset of memory devices are described. In one aspect, a block of data to be written to a line in a rank of memory may be received. The rank of memory may comprise a set of memory devices. The block of data may be compressed. The compressed block of data may be written to a subset of the memory devices that comprise the line. The unwritten portions of the line may not be used to store valid data.