Multi-path Expander Bandwidth Control for CPU Thermal Management
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current storage systems with high data transfer speeds and IOPS requirements can overburden the host CPU, leading to excessive utilization and potential thermal issues, as they provide bandwidth and performance far exceeding typical needs, especially in low-end systems.
Innovation Solution
Implementing a method where a multi-path expander with an embedded processor selectively disables data channel pathways between the controller and storage devices, reducing bandwidth and IOPS performance through software control, allowing for flexible performance tuning and thermal management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If high data transfer speeds and IOPS performance are provided in storage systems, then bandwidth and data rates are improved, but host CPU utilization increases excessively and thermal issues occur
Solution Approach 1:
An expander device is introduced as an intermediary between the host CPU and storage devices. The expander includes an embedded processor that handles data path management, IOPS operations, and bandwidth control, thereby mediating between the CPU and storage devices to reduce CPU utilization while maintaining high performance
Solution Approach 2:
The system dynamically changes operational parameters including bandwidth allocation, data transfer rates, and IOPS performance levels. The embedded processor in the expander adjusts these parameters based on system conditions to optimize performance while controlling CPU utilization and thermal output
2Productivity
If multiple parallel data channel pathways are provided between expander and controller, then bandwidth is improved, but system complexity and cost increase
Solution Approach 1:
The expander's embedded processor dynamically enables or disables data channel pathways based on performance requirements. This allows the system to adapt the number of active pathways from 1 to N, providing high bandwidth when needed while reducing complexity and cost when lower performance is sufficient
3Productivity
If high IOPS performance is provided, then data processing capability is improved, but CPU capacity is consumed at 100% performing IOPS
Solution Approach 1:
The expander's embedded processor acts as an intermediary that handles IOPS operations independently from the host CPU. This offloads IOPS processing to the expander, maintaining high data processing capability while preserving CPU capacity for other computational tasks
Solution Approach 2:
The system segments IOPS handling responsibilities between the expander's embedded processor and the host CPU. The expander manages data path-level IOPS operations while the host CPU handles higher-level storage management, creating a hierarchical division of labor that improves overall system efficiency
Data Source
AI summary
Apparatus and method for controlling the storage of data in a multi-device storage system. In some embodiments, a storage system includes an expander coupled to a controller via multiple parallel data channel pathways, and a plurality of data storage devices coupled to the expander. The expander includes a control circuit adapted to selectively disable at least one but less than all of the multiple parallel data channel pathways.