Configurable Interconnect Resource Allocation for Dynamic Workloads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing data processing systems face inefficiencies due to fixed resource allocation schemes that fail to dynamically adjust to varying requester workloads and traffic type definitions, leading to resource under-utilization and lack of configurability.
Innovation Solution
A configurable multi-tiered interconnect resource allocation system that uses traffic class identifiers and a multi-tier bandwidth management scheme to dynamically allocate resources based on user-defined mappings of transaction attributes, optimizing resource usage across different workload and contention scenarios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a fixed resource allocation scheme is used, then resource allocation is simple to implement, but resource utilization efficiency deteriorates when workload fluctuates
Solution Approach 1:
The patent implements dynamic resource allocation by allowing the resource allocation configuration to be modified at runtime based on workload conditions. The system transitions from a static fixed allocation scheme to a dynamic one where resource capabilities and bandwidth allocations can be adjusted according to actual traffic patterns and contention scenarios, thereby maintaining both ease of implementation through configuration files and high resource utilization efficiency.
Solution Approach 2:
The patent changes the parameters of resource allocation by introducing configurable resource capability data that defines bandwidth allocations for different traffic classes. These parameters can be modified without changing the underlying system architecture, allowing the same interconnect to operate under different allocation schemes (e.g., equal bandwidth vs. unequal bandwidth) based on workload requirements.
2Device complexity
If a fixed resource allocation scheme is used, then system complexity is reduced, but adaptability to different workload scenarios deteriorates
Solution Approach 1:
The system achieves adaptability through dynamic configuration rather than dynamic hardware reconfiguration. The resource allocation scheme can be changed by loading different configuration data, allowing the system to adapt to various workload scenarios (e.g., voice traffic prioritization, data traffic balancing) without increasing physical device complexity.
Solution Approach 2:
The patent uses parameter-based configuration to provide adaptability. By defining traffic class mappings, resource capability data, and bandwidth allocations as configurable parameters, the system can adapt to different scenarios simply by changing these parameters rather than modifying the system architecture or adding complex hardware logic.
3Ease of operation
If equal bandwidth is allocated to all traffic classes, then fairness is improved, but overall system performance deteriorates when contention occurs
Solution Approach 1:
The patent applies local quality by allocating different bandwidth resources to different traffic classes based on their specific requirements. Instead of uniform allocation, critical traffic classes (e.g., real-time voice) receive higher bandwidth allocations during contention, while less critical classes receive lower allocations, optimizing overall system performance while maintaining operational simplicity through automated classification.
Solution Approach 2:
The system dynamically adjusts bandwidth allocation based on traffic class and contention conditions. During non-contention periods, equal bandwidth may be provided for fairness, but during contention scenarios, the system automatically redistributes bandwidth according to configured priorities and resource capability data, thereby improving system performance without sacrificing fairness when conditions permit.
Data Source
AI summary
The present disclosure advantageously provides a method and system for allocating shared resources for an interconnect. A request is received at a home node from a request node over an interconnect, where the request represents a beginning of a transaction with a resource in communication with the home node, and the request has a traffic class defined by a user-configurable mapping based on one or more transaction attributes. The traffic class of the request is determined. A resource capability for the traffic class is determined based on user configurable traffic class-based resource capability data. Whether a home node transaction table has an available entry for the request is determined based on the resource capability for the traffic class.