Cloud Resource Instruction Set Architecture for Dynamic Workload Adaptation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Cloud computing environments face challenges such as fast-changing system configuration requirements due to workload constraints, varying innovation cycles of system components, and the paradox between maximal performance and maximal sharing of systems and subsystems, leading to inefficiencies in traditional data center designs.

Innovation Solution

A cloud resource instruction set architecture (CRISA) is developed, utilizing a high-throughput, low-latency network as a system backplane, composable system building blocks, and a multidimensional capability and capacity relationship abstraction to enable resource optimization and self-tuning, allowing for dynamic resource allocation and maximal sharing across subsystems.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If traditional data center design is used, then system stability is maintained, but adaptability to fast-changing workload constraints deteriorates

Engineering Contradiction:

Improveadaptability to workload constraintsVSAvoidsystem configuration complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system is divided into independent composable building blocks (compute nodes, storage nodes, network nodes) that can be individually configured and assembled. Each node type is a self-contained unit with standardized interfaces, allowing rapid reconfiguration of the overall system to match changing workload requirements without redesigning the entire data center.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system configuration is made dynamic through software-defined networking and virtualization layers that allow real-time adjustment of resource allocation and network topology. This enables the system to adapt to varying workload constraints by dynamically provisioning resources rather than requiring static physical reconfiguration.

Inventive Principle:

Principle #15Dynamics

2Productivity

If system components are updated to follow innovation cycles, then performance is improved, but system stability deteriorates

Engineering Contradiction:

Improvesystem performanceVSAvoidsystem stability

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

Components are segmented into modular units with standardized interfaces and protocols. This allows individual components to be updated or upgraded independently without affecting the entire system, enabling selective adoption of innovative components while maintaining stability of proven components.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

Universal interfaces and communication protocols are established across all component types, allowing new components with innovative features to be integrated into the existing system without requiring system-wide redesign. The standardized interfaces ensure that innovations can be adopted selectively while maintaining overall system stability.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Productivity

If systems are maximally shared to improve resource utilization, then resource efficiency is improved, but performance deteriorates due to contention

Engineering Contradiction:

Improveresource utilization efficiencyVSAvoidsystem throughput

Core Design Contradiction:

ProductivityVSSpeed

Solution Approach 1:

A software-defined networking layer acts as an intermediary between physical resources and workload demands. This virtualization layer abstracts resource sharing mechanics, enabling multiple workloads to access shared resources efficiently through intelligent scheduling and resource allocation algorithms that minimize contention while maximizing utilization.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

Resource allocation parameters such as bandwidth, latency thresholds, and priority levels are dynamically adjusted based on current system state and workload characteristics. This allows the system to optimize the balance between resource sharing and performance by changing operational parameters rather than physical configuration.

Inventive Principle:

Principle #35Parameter changes

4Adaptability or versatility

If network is configured as backplane to enable maximal sharing, then resource sharing is improved, but latency increases

Engineering Contradiction:

Improveresource sharing capabilityVSAvoidnetwork latency

Core Design Contradiction:

Adaptability or versatilityVSLoss of time

Solution Approach 1:

Software-defined networking introduces an intelligent intermediary layer that manages network traffic between shared resources. This layer implements sophisticated routing, quality of service policies, and traffic prioritization that reduce latency for time-sensitive operations while maintaining high resource sharing capability through efficient multiplexing.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS10432464B2Creating new cloud resource instruction set architecture

Publication Date: 2019.10.01 INTERNATIONAL BUSINESS MACHINE CORPORATION

AI summary

A method and system are provided. The system includes a network configurator for configuring a network as a backplane of the system to optimize throughput rate and minimize latency across a plurality of subsystems that are cloud-based and that form the system. The system further includes a composable system building block configurator for refactoring the plurality of subsystems to use the network to communicate as a single system. The system also includes a system resource multidimensional model generator for generating a multidimensional optimization model that models the composable system building blocks as resources having adjustable parameters in a multidimensional parameter space.