Distributed Compute Clusters with Complementary Energy Scheduling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cloud computing systems face challenges in utilizing renewable energy due to the high variability of power production from sources like solar and wind, which can lead to intermittent availability of computing resources, causing downtime and carbon footprint issues.
Innovation Solution
A distributed compute platform with multiple compute clusters located near renewable energy sources, each equipped with a local scheduler that adjusts computational demand to match energy availability, using a global scheduler to distribute virtual machines across clusters with complementary energy patterns, thereby stabilizing compute resources and reducing carbon footprint.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-generated harmful factors
If cloud computing providers use renewable energy sources to power data centers, then carbon footprint is reduced, but service performance degrades due to high variability of power production
Solution Approach 1:
The patent segments the cloud computing infrastructure into multiple geographically distributed compute clusters, each co-located with different renewable energy sources. This segmentation allows the system to leverage diverse energy patterns across locations while maintaining service reliability through distributed architecture.
Solution Approach 2:
The patent merges multiple renewable energy sources with complementary production patterns into a unified power supply system for the distributed compute clusters. By combining solar, wind, and other renewable sources across different locations, the system achieves more stable aggregate power availability while maintaining carbon neutrality.
2Object-generated harmful factors
If compute clusters are powered by local renewable energy sources with variable patterns, then carbon footprint is reduced, but compute resource availability becomes intermittent causing downtime
Solution Approach 1:
The patent implements dynamic workload placement and migration mechanisms that adapt compute task distribution in real-time based on renewable energy availability at each compute cluster. The system dynamically shifts workloads between clusters according to varying energy patterns to maintain continuous service availability.
Solution Approach 2:
The patent introduces an energy management intermediary layer that coordinates between renewable energy sources and compute clusters. This intermediary system predicts energy availability, pre-positions workloads, and manages workload migration to bridge gaps in renewable energy production and maintain service continuity.
3Reliability
If a distributed compute platform uses multiple compute clusters with complementary energy patterns, then compute resource stability is improved, but system complexity increases
Solution Approach 1:
The patent designs compute clusters with multi-functional capabilities that can operate independently or in coordination with other clusters. Each cluster is designed to handle a universal set of workloads while also being optimized for its local energy source characteristics, enabling flexible deployment and simplified management.
Solution Approach 2:
The patent implements feedback mechanisms that continuously monitor energy availability, compute workload status, and system performance across distributed clusters. This feedback information is used to automatically adjust workload distribution, optimize resource allocation, and maintain system stability without requiring complex manual intervention.
Data Source
AI summary
A computer system that includes a plurality of compute clusters that are located at different geographical locations. Each compute cluster is powered by a local energy source at a geographical location of that compute cluster. Each local energy source has a pattern of energy supply that is variable over time based on an environmental factor. The computer system further includes a server system that executes a global scheduler that distributes virtual machines that perform compute tasks for server-executed software programs to the plurality of compute clusters of the distributed compute platform. To distribute virtual machines for a target server-executed software program, the global scheduler is configured to select a subset of compute clusters that have different complementary patterns of energy supply such that the subset of compute clusters aggregately provide a target compute resource availability for virtual machines for the target server-executed software program.