Storage Compute Device Power Management via Section Deactivation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional storage devices are inefficient in managing power consumption during computations, particularly for applications like neuromorphic computing and scientific simulations, where large matrices are processed in parallel, leading to bottlenecks in bus lines and I/O protocols and increased power usage.
Innovation Solution
A storage compute device with multiple data storage sections that facilitate parallel read/write operations, where at least one section can be deactivated during computations to reduce power usage, utilizing a processing unit to manage power consumption based on parametric information and performance needs, and employing selective power management to balance power and performance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If multiple data storage sections are activated to handle parallel read/write operations for computations, then processing capability and productivity are improved, but power consumption increases
Solution Approach 1:
The system dynamically activates or deactivates storage sections based on real-time computational demands. The processing unit monitors workload requirements and adjusts the operational state of storage sections accordingly, transitioning from static to dynamic power management to optimize the balance between processing capability and power consumption.
Solution Approach 2:
Instead of keeping all storage sections continuously activated, the system activates only the necessary portion required for current computational tasks. This partial action approach ensures that sufficient processing capability is maintained while avoiding the excessive power consumption that would result from keeping all sections active.
2Reliability
If all storage sections remain active to ensure data availability, then reliability is improved, but power usage increases
Solution Approach 1:
The system performs preliminary assessment of data access requirements before deactivating storage sections. By predicting which sections will be needed for upcoming operations, the system can maintain data availability for anticipated tasks while deactivating sections that are not immediately required, thus reducing power usage without compromising reliability.
Solution Approach 2:
The processing unit continuously monitors system state and provides feedback to control the activation status of storage sections. This closed-loop control ensures that sections are deactivated only when it is safe to do so, maintaining data availability while optimizing power consumption based on real-time system conditions.
Data Source
AI summary
Computations are performed on data objects via two or more data storage sections. The data storage sections facilitate persistently storing the data objects in parallel read/write operations. The data objects are used in computations within a storage compute device. At least one of the storage sections is deactivated during the computations to reduce power usage of the storage compute device.