Flattened Data Storage Hierarchy Reducing Latency and Energy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data center architectures face inefficiencies due to complex memory hierarchies, high energy consumption, and scalability issues, with excessive material usage, latency, and reliability concerns related to multiple memory levels and mechanical hard disks.
Innovation Solution
Implementing a flattened-level data storage hierarchy using solid-state technologies like memristor, PCRAM, or flash memory, which eliminates the need for separate main memory and long-term data storage areas, and incorporates a local processor for autonomous operation and reduced cache layers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If a complex memory hierarchy with multiple cache layers and separate main memory is used, then data access performance is improved, but device complexity and material usage increase
Solution Approach 1:
The patent combines main memory and long-term data storage areas into a single unified non-volatile data store, eliminating the need for separate memory components and complex hierarchy management. This merging reduces device complexity while maintaining data access performance through direct processor access to the unified storage medium.
Solution Approach 2:
The non-volatile data store serves multiple functions simultaneously - acting as both main memory for frequent access and long-term storage for persistent data, replacing the need for separate cache layers and memory components. This multi-functionality simplifies the overall system architecture while maintaining performance.
2Quantity of substance
If multiple memory hierarchy levels are implemented, then data storage capacity is increased, but energy consumption increases
Solution Approach 1:
By merging main memory and long-term storage into a single non-volatile data store, the patent eliminates the energy required to maintain multiple separate memory hierarchy levels. The unified structure provides both frequent and long-term storage capabilities without the continuous power consumption associated with traditional multi-level memory systems.
3Reliability
If separate main memory and long-term data storage areas are used, then data storage reliability is improved, but material usage and device complexity increase
Solution Approach 1:
The patent merges separate memory and storage components into a single non-volatile data store, reducing material usage while maintaining reliability. The non-volatile nature of the unified storage ensures data persistence and reliability without requiring duplicate storage areas or complex redundancy mechanisms.
4Quantity of substance
If traditional data center architectures with mechanical hard disks are used, then data storage capacity is increased, but latency and reliability concerns increase
Solution Approach 1:
The patent replaces mechanical hard disk systems with solid-state non-volatile memory technology, eliminating mechanical moving parts. This substitution dramatically reduces data access latency while maintaining high storage capacity, and improves reliability by removing mechanical failure points.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Data storage apparatus and methods are disclosed. A disclosed example data storage apparatus comprises a cache layer and a processor in communication with the cache layer. The processor is to dynamically enable or disable the cache layer via a cache layer enable line based on a data store access type.