Cloud Firmware vBIOS for Scalable Core Partitioning

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current cloud computing architectures face challenges with scalability, latency, and complexity due to the need for OS and virtualization layers, which degrade as the number of processor cores increases, and introduce inefficiencies in resource management.

Innovation Solution

A cloud firmware using virtual BIOS (vBIOS) is dynamically created to partition and manage hardware resources without requiring an operating system or virtual machine manager, allowing for efficient and flexible resource allocation across multiple processor cores.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If OS and virtualization layers are used to manage cloud computing resources, then resource management capability is improved, but system complexity and latency increase

Engineering Contradiction:
Improveresource management capabilityVSAvoidsystem complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent extracts and removes the OS and virtualization layers from the cloud computing architecture, retaining only the essential firmware layer (vBIOS) for resource management. This extraction eliminates unnecessary complexity while preserving core resource allocation capabilities through direct hardware-firmware interaction.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The vBIOS serves as an intermediary layer between hardware and applications, providing essential resource management functions without requiring full OS or virtualization stacks. This mediator enables direct hardware access and management while reducing system complexity.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Adaptability or versatility

If OS and virtualization layers are used to manage cloud computing resources, then resource management capability is improved, but latency increases

Engineering Contradiction:
Improveresource management capabilityVSAvoidlatency
Core Design Contradiction:
Adaptability or versatilityVSLoss of time

Solution Approach 1:

By removing the OS and virtualization layers, the patent eliminates the time-consuming operations associated with these layers, such as context switching, system call processing, and virtualization overhead. Applications can execute directly on hardware through vBIOS-managed resources, significantly reducing latency.

Inventive Principle:
Principle #2Taking out (Extraction)

3Productivity

If the number of processor cores increases to improve scalability, then computing capacity is improved, but OS and virtualization layers degrade in performance

Engineering Contradiction:
Improvecomputing capacityVSAvoidOS and virtualization layer performance degradation
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent segments the system into two distinct layers: vBIOS for hardware resource management and applications for computational tasks. This segmentation allows each layer to operate independently and optimally, with the vBIOS efficiently managing resources across multiple cores without the performance degradation that plagues OS and virtualization layers in many-core environments.

Inventive Principle:
Principle #1Segmentation

4Adaptability or versatility

If vBIOS dynamically partitions processor cores without restarting CPU, then resource allocation flexibility is improved, but system stability may be affected

Engineering Contradiction:
Improveresource allocation flexibilityVSAvoidsystem stability
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The vBIOS implements dynamic resource partitioning that allows processor cores to be allocated and reallocated on-the-fly without system restart. This dynamic capability enables flexible resource management while maintaining system stability through careful implementation that ensures proper resource isolation and assignment during transitions.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP3161621B1Cloud firmware
Publication Date: 2020.10.07 HUAWEI TECH CO LTD
  • EP3161621B1 patent drawingFigure 1~2
  • EP3161621B1 patent drawingFigure 3~4
  • EP3161621B1 patent drawingFigure 5

AI summary

A network element (NE) comprising a receiver configured to couple to a cloud network; and a multi-core central processing unit (CPU) coupled to the receiver and configured to receive a first partition configuration from an orchestration element, partition a plurality of processor cores into a plurality of processor core partitions according to the first partition configuration, and initiate a plurality of virtual basic input/output systems (vBIOSs) such that each vBIOS manages a processor core partition.