Configurable Processing Units with Multicast Cache Routing

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current data processing systems face challenges in providing independent and efficient graphics processing operations for multiple isolated sub-systems, such as those found in vehicles, where time-sharing a single GPU or dedicating separate GPUs for each function is either insufficient or costly.

Innovation Solution

A data processing system with configurable processing units that can be partitioned into master-slave configurations, utilizing a multicast communication network to route cache communications efficiently across processing units, allowing for dynamic allocation and reconfiguration of resources based on current requirements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If a single graphics processing unit is time-shared between different graphics processing functions, then resource utilization is improved, but independence and isolation between sub-systems deteriorates

Engineering Contradiction:

Improveresource utilizationVSAvoidindependence and isolation between sub-systems

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The processing units are divided into multiple partitions, where each partition can be independently allocated to different sub-systems. This segmentation allows multiple isolated graphics processing functions to operate simultaneously on different partitions while sharing the same physical processing units, thus maintaining both resource utilization and sub-system independence.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The system provides dynamic configurability where processing units can be reconfigured between partitions at runtime based on current requirements. This dynamic allocation allows the system to adapt resource distribution while maintaining isolation boundaries, resolving the contradiction between fixed resource allocation for independence and flexible allocation for utilization.

Inventive Principle:

Principle #15Dynamics

2Reliability

If completely separate graphics processing units are provided for each graphics processing function, then independence and isolation between sub-systems is improved, but device complexity and cost deteriorates

Engineering Contradiction:

Improveindependence and isolation between sub-systemsVSAvoidnumber of processing components

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

Multiple processing units are merged into a single system that can be partitioned and allocated to different sub-systems. By combining multiple processing units into one configurable system with partitioning capabilities, the invention achieves the isolation of separate units while reducing the overall complexity and cost of having completely separate physical devices for each function.

Inventive Principle:

Principle #5Merging (Combining)

Solution Approach 2:

The processing units are designed to be universal and multi-functional, capable of serving different graphics processing functions through configurable partitions. This universality allows a single set of processing units to replace multiple dedicated units, reducing device complexity while maintaining the independence required for different sub-systems through software-defined partitioning.

Inventive Principle:

Principle #6Universality (Multi-functionality)

3Adaptability or versatility

If processing units are dynamically partitioned and reconfigured, then adaptability is improved, but device complexity deteriorates

Engineering Contradiction:

Improveflexibility in resource allocationVSAvoidconfiguration management complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

A configuration system acts as an intermediary between the physical processing units and the logical partitions. This intermediary layer manages the complexity of dynamic reconfiguration by providing abstraction and control mechanisms, allowing flexible resource allocation without exposing the full complexity of the reconfiguration process to external systems.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS11489940B2Data processing systems

Publication Date: 2022.11.01 ARM LTD

AI summary

A data processing system comprising a plurality of processing units that are configurable as different partitions of processing units. The system comprises a multicast communication network for routing cache communications within a partition of processing unit. A cache controller of one of the processing units within a partition is configurable as a master cache controller for a set of caches within the partition. The master cache controller is operable to issue requests over the multicast communication network to all of the caches in the set of caches at the same time. The multicast communication network is configured to combine response signals from the different processing units within the partition to provide a combined response signal to the master cache controller that represents an overall request-response status for the caches to which the request was issued.