Composable Core Matrix for Service-Level Compliance and Power Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data storage systems consume significant electrical power due to server operation and cooling needs, with existing systems often running all processor cores at maximum clock speed, leading to inefficient power usage and potential performance violations.
Innovation Solution
A storage system is configured with processor core pools corresponding to different service levels, allowing cores to operate at defined clock speeds matching the service level requirements, and a core matrix controller dynamically adjusts core allocations based on forecasted IOPS demand to maintain compliance and reduce power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If all processor cores run at maximum clock speed, then processing speed and productivity are improved, but power consumption increases significantly
Solution Approach 1:
The processor cores are segmented into multiple pools (e.g., first pool, second pool, third pool) where each pool operates at different clock speeds. This segmentation allows the system to divide workloads across cores with varying performance levels, achieving a balance between processing speed and power consumption without requiring all cores to run at maximum speed.
Solution Approach 2:
Different pools of processor cores are assigned different clock speeds tailored to their specific functional requirements. For example, pools handling time-sensitive operations run at higher clock speeds while pools handling less time-critical tasks run at lower clock speeds. This local quality approach optimizes the trade-off between performance and energy consumption for each core pool individually.
2Use of energy by moving object
If processor clock speed is reduced to save power, then power consumption decreases, but service level compliance may be violated
Solution Approach 1:
The system dynamically adjusts the clock speeds of different processor pools based on real-time workload characteristics and service level requirements. When workloads demand higher performance, the system increases clock speeds for relevant pools; when workloads are lighter, it reduces clock speeds to conserve power. This dynamic adjustment ensures service level compliance is maintained while optimizing power consumption.
Solution Approach 2:
The system changes the operational parameters (clock speeds) of processor pools based on workload conditions and service level agreements. By adjusting clock speeds as a variable parameter rather than maintaining a fixed high speed, the system can adapt to varying demand patterns, ensuring compliance with service levels while minimizing power consumption during lower-demand periods.
3Device complexity
If a single uniform clock speed is used for all processor pools, then system simplicity is maintained, but power consumption efficiency deteriorates
Solution Approach 1:
The processor cores are divided into multiple pools that can operate independently at different clock speeds. This segmentation introduces manageable complexity in the form of pooled architecture, which enables differentiated performance levels and power consumption characteristics across pools without requiring complete system redesign.
Solution Approach 2:
Rather than implementing a completely uniform clock speed across all cores, the system applies partial differentiation by implementing multiple clock speed levels for different pools. This partial action approach achieves power consumption efficiency improvements without the excessive complexity of fully customized per-core frequency control, striking a practical balance between simplicity and energy efficiency.
Data Source
AI summary
A storage system is configured with pools of processor cores. Each pool corresponds uniquely to one of the supported service levels of the storage system. Processor cores within each pool run at a clock speed that is defined for the service level corresponding to the respective pool. Incoming IOs are enqueued for processing by cores of the pool corresponding to the service level of the storage group containing the storage object that is the target of the IO. Forecast IO demand and corresponding data access latency are computed for each service level. If the forecast latency of a service level includes values outside a compliance range, then cores are promoted to, or demoted from, the corresponding pool so that forecast latency stays within the compliance range, thereby reducing power consumption while promoting service level compliance.