Virtual Motherboards and Storage via Node Disaggregation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional data centers face inefficiencies in resource allocation, leading to underutilization of computing resources, high power consumption, and complex maintenance due to the allocation of computational and storage resources in units of physical server devices, which cannot accommodate granular requests from clients and result in wasted resources and increased costs.
Innovation Solution
A computing system architecture that disaggregates computing nodes and storage nodes, allowing them to be remotely situated and connected via high-bandwidth networks, enabling the creation of virtual motherboards and storage devices at runtime, which can be allocated based on client requests, optimizing resource utilization and simplifying maintenance.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If computational and storage resources are allocated in units of physical server devices, then resource allocation is simplified, but resource utilization efficiency deteriorates due to inability to accommodate granular client requests
Solution Approach 1:
The patent segments physical server devices into smaller computational units (virtual machines, containers, or microservices) that can be independently allocated. This allows cloud operators to provide granular resources to clients while maintaining simplified management at the physical device level, resolving the contradiction between allocation simplicity and utilization efficiency.
Solution Approach 2:
The patent implements dynamic resource allocation where computational units can be flexibly created, modified, and destroyed based on client requests. This dynamic approach enables precise matching of resources to needs, improving utilization efficiency while maintaining operational simplicity through automated provisioning.
2Ease of operation
If entire server devices are allocated to clients, then resource allocation is straightforward, but computing resources are under-utilized when clients request only partial capacity
Solution Approach 1:
The patent merges multiple physical server devices into a pooled resource pool that can be dynamically shared among multiple clients. By combining resources at the physical level and allocating virtually, the system achieves both straightforward allocation (through a unified pool) and high utilization (through shared access), eliminating the waste of allocating entire servers when only partial capacity is needed.
3Reliability
If conventional physical server devices are used, then hardware functionality is complete, but power consumption and cooling costs increase due to large CPU power requirements
Solution Approach 1:
The patent creates virtual copies of computing resources that can be instantiated and destroyed on demand. Instead of maintaining permanent physical server devices with full hardware functionality, the system uses virtualization to create lightweight representations that consume minimal power, providing complete functionality only when needed and eliminating continuous power consumption of idle hardware.
4Reliability
If individual components on motherboards are replaced during maintenance, then system reliability is maintained, but maintenance complexity and cost increase due to multiple disassembly steps
Solution Approach 1:
The patent replaces the mechanical hardware maintenance model with a virtualized software-based system. Instead of physically disassembling server devices to replace motherboard components, the system uses virtual machine migration and software updates to maintain and update computing resources, eliminating the need for complex physical repairs while maintaining system reliability.
Data Source
AI summary
Described herein are various technologies pertaining to a computing system architecture that facilitates construction of virtual motherboards and virtual storage devices. A computing system includes a plurality of computing nodes and a plurality of storage nodes, where the computing nodes are disaggregated from the storage nodes. The computing nodes include respective system on chip modules. The computing nodes and the storage nodes are communicatively coupled by way of a full bisection bandwidth network, where each storage node is allocated network bandwidth that at least matches the maximum input/output speed of the storage node. Responsive to receipt of a client request, computing nodes are allocated to the client and exposed to an application of the client as a motherboard. Likewise, storage nodes are allocated to the client and exposed to the application of the client as a larger storage device.


