Adaptive Workload Framework for Computing Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computing networks require manual hardcoding and library modifications to route workloads effectively across changing computing units, such as CPUs and APUs, which is inefficient and requires user expertise, especially when new APUs or CPU functionalities are added.
Innovation Solution
A software framework that automatically detects changes in computing units and selects the optimal units for workload execution based on data parameters, eliminating the need for manual hardcoding and providing uniform APIs for workload execution across various functionalities like encryption, decryption, and AI functions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If manual hardcoding and library modifications are used to route workloads to computing units, then workload routing can be customized for specific configurations, but the system complexity increases and requires user expertise when computing units change
Solution Approach 1:
The patent introduces an intermediary layer (the framework) between the workload submission interface and the physical computing units. This framework automatically detects computing unit changes, maintains a mapping between workload types and available computing units, and handles the routing logic without requiring library modifications. The intermediary absorbs the complexity of adapting to hardware changes while presenting a stable interface to users.
Solution Approach 2:
The system implements self-service by automatically detecting when computing units are added or removed and autonomously updating its internal mappings and routing logic. The framework monitors the computing environment, identifies changes in available APUs or CPU functionalities, and adjusts workload distribution without human intervention, eliminating the need for manual hardcoding updates.
2Productivity
If computing units are frequently changed to improve performance, then computing capability increases, but the ease of operation decreases due to required library modifications
Solution Approach 1:
The framework automatically adapts to computing unit changes by detecting new APUs or CPU functionalities and updating its internal state without requiring user action. This self-service mechanism maintains ease of operation even as computing capability evolves, as users simply submit workloads through the standard interface while the framework handles all adaptation logic autonomously.
Solution Approach 2:
The system implements dynamic adaptation where the framework continuously monitors the computing environment and adjusts its workload routing strategy in real-time based on available resources. This dynamic behavior allows the system to optimize for current computing capabilities while maintaining a stable user interface, resolving the contradiction between improving productivity through hardware changes and maintaining ease of operation.
3Ease of operation
If uniform APIs are implemented across different computing units, then ease of operation improves, but the ability to optimize for specific unit characteristics may be reduced
Solution Approach 1:
The framework serves as an intelligent intermediary that receives uniform API calls from users and translates them into optimized routing decisions based on the specific characteristics of available computing units. The uniform interface maintains ease of operation, while the intermediary layer preserves optimization capabilities by matching workload requirements with the most suitable computing units using its knowledge of unit characteristics and performance metrics.
Solution Approach 2:
The framework implements a universal interface that can handle multiple types of workloads across different computing unit types through a single unified API. This universal approach maintains ease of operation while the underlying framework maintains detailed knowledge of specific unit characteristics, enabling it to optimize routing decisions without requiring users to interact with multiple specialized interfaces.
Data Source
AI summary
A processing circuitry, a method to be performed at the processing circuitry, a computer-readable storage medium, and a computing system. The processing circuitry is to determine a first mapping between a first set of data parameters and first computing units of a computing network; select, based on the first mapping and on first data having a first workload associated therewith, one or more of the first computing units to execute the first workload, and send for execution the first workload to the one or more of the first computing units; determine a second mapping based on a change in computing units from the first computing units to second computing units, the second mapping between a second set of data parameters and the second computing units; and select, based on the second mapping and on second data having a second workload associated therewith, one or more of the second computing units to execute the second workload.