Virtual Machine and Hardware IP Block for Low Power Sensing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Constrained cost and power budgets in computing devices, such as smartphones, make it challenging to implement dedicated hardware for all platform sensing requirements, while hardware-less solutions fail to meet industry standards for certain use cases, necessitating a cost-efficient and power-efficient sensor hub solution.
Innovation Solution
A low-power sensing scheme is achieved by splitting sensing functionality between a virtual machine (VM) and a small dedicated hardware (HW) Intellectual Property (IP) block, where the VM provides sufficient memory and compute capabilities, and the HW IP block offers a cost-effective sensing solution for low-power use cases, maintaining a unified interface with the operating system to hide hardware limitations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If dedicated hardware sensor hub is used, then sensing capability and reliability are improved, but power consumption and cost increase
Solution Approach 1:
The patent segments the sensing system into two parts: a minimal dedicated hardware IP block for basic sensing operations and a virtual machine for complex processing. This segmentation allows the hardware to be simplified (reducing power and cost) while the virtual machine handles advanced sensing tasks that require higher processing capability, thus maintaining reliability without the need for a full-featured hardware sensor hub.
Solution Approach 2:
The virtual machine acts as an intermediary between the simplified hardware IP block and the operating system. It translates high-level sensing requirements into hardware-friendly operations, enabling the system to achieve sophisticated sensing capabilities through software rather than requiring equivalent hardware complexity, thereby reducing power consumption and cost.
2Reliability
If dedicated hardware sensor hub is used, then sensing capability is improved, but device cost increases
Solution Approach 1:
The sensing system is divided into a minimal hardware IP block and a software-based virtual machine. This segmentation reduces the hardware resource requirements (memory, computation units) in the dedicated IP block, directly lowering manufacturing costs while the virtual machine provides the necessary processing power through software execution on existing processor resources.
Solution Approach 2:
Instead of duplicating full sensor hub hardware for each sensing task, the patent uses a virtual machine that can be instantiated and configured through software to handle different sensing workloads. This software-based copying approach is more cost-effective than hardware duplication, as it leverages the existing processor's capabilities rather than requiring additional dedicated hardware components.
3Ease of manufacture
If hardware-less sensing solution is used, then cost is reduced, but power consumption increases
Solution Approach 1:
The patent introduces a segmented architecture where a minimal dedicated hardware IP block handles basic sensing operations that benefit from hardware acceleration (lower power consumption), while the virtual machine handles more complex processing tasks. This segmentation ensures that cost-effective hardware is used for appropriate functions, preventing the power consumption issues that would arise from a completely software-based approach.
Solution Approach 2:
The dedicated hardware IP block is designed to autonomously perform basic sensing operations without requiring full processor activation. It can independently collect and pre-process sensor data, only engaging the higher-power virtual machine when necessary, thus achieving low power consumption while maintaining cost-effectiveness through selective hardware utilization.
Data Source
AI summary
Described is an apparatus comprising: a processor operable to execute a virtual machine manager (VMM) which is to manage a virtual machine (VM) for a hardware intellectual property (IP) block; a communication fabric; and a hardware IP block coupled to the processor via the communication fabric, wherein the hardware IP block is to be coupled to a first set of one or more sensors, and wherein the VM and the hardware IP block are operable to process data collected from the first set.


