Cooperative Storage Architecture for SSD Interconnect
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The increasing complexity and congestion in data communication and signal transmission between multiple solid state drives (SSDs) and a host computing device in storage appliances lead to significant computational and memory bandwidth loads, limiting the growth of storage capacity and efficiency.
Innovation Solution
An interconnect architecture that enables direct communication and data transmission among SSDs through a cooperative flash translation layer (FTL) and buffer managers, allowing device-to-device communication without involving the host computing device, thereby offloading computational and memory bandwidth loads.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multiple solid state drives are utilized to construct a large data storage appliance, then storage capacity increases, but computational load and memory bandwidth load on the host computing device significantly increase
Solution Approach 1:
The patent divides the computational workload by creating autonomous SSD agents that can independently process commands and make decisions about data placement and retrieval. Each SSD contains its own controller with processing capabilities, allowing the system to segment the centralized host computing responsibilities into distributed edge computing functions at the SSD level. This segmentation enables storage capacity to scale without proportionally increasing host computational load.
Solution Approach 2:
The patent introduces an inter-SSD communication protocol and message passing mechanism as an intermediary layer between SSDs and the host. This intermediary enables direct peer-to-peer communication and coordination among SSDs, allowing them to share workload and coordinate data operations without constantly involving the host computing device. The intermediary protocol handles command routing, data placement decisions, and wear-leveling coordination autonomously.
2Quantity of substance
If multiple solid state drives are utilized to construct a large data storage appliance, then storage capacity increases, but memory bandwidth load on the host computing device significantly increase
Solution Approach 1:
The patent extracts memory-intensive operations from the host computing device and relocates them to the SSD controllers. Each SSD controller maintains local metadata, allocation tables, and wear-leveling information in its own memory, eliminating the need for the host to manage these data structures. This extraction of memory management functions to the edge devices reduces memory bandwidth requirements at the host while enabling storage capacity to scale.
Solution Approach 2:
The patent changes the communication dimension by enabling direct SSD-to-SSD interactions bypassing the host computing device. The inter-SSD protocol allows SSDs to exchange commands, data, and coordination messages directly through dedicated communication channels, creating a parallel communication dimension that does not consume host memory bandwidth. This dimensional change enables storage expansion without proportionally increasing host memory bandwidth requirements.
3Adaptability or versatility
If the host computing device reaches its maximum computational load, then the number of solid state drives allowed under the host reaches its limit, but this constrains further growth of the storage appliance capacity
Solution Approach 1:
The patent introduces dynamic autonomy to SSD controllers, enabling them to adaptively manage their own operations and coordinate with other SSDs based on real-time system conditions. The SSD agents can dynamically decide whether to execute commands locally or forward them to the host, dynamically adjust wear-leveling strategies, and dynamically allocate data placement based on available capacity and performance conditions. This dynamic behavior allows the storage appliance to grow beyond host computational limits.
Solution Approach 2:
The patent implements self-service capabilities in SSD controllers through autonomous agents that can independently process read/write commands, manage their own wear-leveling, handle error correction, and coordinate data operations with other SSDs without host intervention. The SSDs serve themselves by maintaining local state information, making autonomous decisions about data placement and retrieval, and self-managing their operational parameters, thereby freeing the host from managing individual drive operations.
Data Source
AI summary
The present disclosure provides an interconnect architecture that enables communications and/or data transmissions among data storage drives in a computing system. The flash translation layer (FTL) in each data storage drive may be operated in a cooperative manner that allows communications and/or data transmissions across memory arrays from each of the data storage drives implemented in the computing system. The direct communications and/or data transmissions among the data storage drives in the computing system may be enabled without deferring back to a host computing device in the computing system. Thus, the computational load to the host computing device is reduced and the flexibility of scaling up the storage appliance in the computing system is increased.