Parallel Computing Nodes with Autonomous Domain Networking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current computer systems based on CPUs are inadequate for meeting the demands of massive data and large-scale computing power required by artificial intelligence and blockchain due to limitations in word length accuracy and serial processing, as well as inefficient network architectures that waste computing power and retransmit data excessively.
Innovation Solution
A computer system comprising multiple computing devices connected in parallel, forming a network topology with autonomous domains, where each device includes storage and computing units integrated into the network, allowing for parallelized computing and storage, and utilizing a communication device with autonomous domain building and task processing modules to efficiently manage resource allocation and data processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If CPU-based serial processing is used, then system reliability is maintained, but data processing speed and computing power are insufficient for massive data demands
Solution Approach 1:
The patent segments the monolithic CPU architecture into multiple independent computing devices, each with its own storage and computing units. These devices are organized into nodes that form a network topology, allowing parallel processing of data while maintaining individual device reliability. Each node can independently process data packets, eliminating the serial processing bottleneck of traditional CPUs.
Solution Approach 2:
The patent transitions from a single-dimension serial processing architecture to a multi-dimensional parallel architecture by introducing network topology with multiple nodes operating simultaneously. Data can flow through multiple paths and be processed by multiple nodes at the same time, adding spatial dimensionality to the processing architecture and dramatically increasing throughput.
2Productivity
If traditional network architecture with protocol operations is used, then compatibility is maintained, but computing power is wasted on protocol operations rather than data processing
Solution Approach 1:
The patent extracts and removes traditional protocol operation layers from the data processing path. Instead of requiring full protocol stacks at each node, the system uses simplified communication mechanisms where nodes can directly exchange data packets with minimal processing overhead. This extraction eliminates the computing power waste associated with protocol operations while maintaining network functionality.
Solution Approach 2:
Each computing device in the patent is self-sufficient with integrated storage and computing units, allowing it to independently process data packets without requiring complex coordination or protocol operations with other nodes. The autonomous domain building capability enables nodes to self-organize and self-manage their processing tasks, reducing the need for centralized control and protocol overhead.
3Quantity of substance
If bandwidth resources are increased to meet AI and blockchain demands, then data transmission capacity is improved, but network complexity and resource allocation difficulty increase
Solution Approach 1:
The patent implements dynamic resource allocation where computing devices can be dynamically added or removed from the network, and autonomous domains can be dynamically built or reconfigured based on current processing demands. This dynamic architecture allows the network to adapt to varying data transmission requirements without requiring permanent over-provisioning of resources, managing complexity through flexibility rather than rigid structure.
Solution Approach 2:
Each computing device in the patent is designed as a multi-functional unit that can perform storage, computing, and communication tasks. This universality allows any node to fulfill multiple roles depending on demand, eliminating the need for specialized high-bandwidth infrastructure and reducing overall network complexity while maintaining high data transmission capacity.
4Productivity
If parallel computing devices are integrated into network topology, then data processing speed is improved, but communication overhead and coordination difficulty increase
Solution Approach 1:
The patent implements preliminary action through autonomous domain building, where nodes are pre-organized into domains based on their capabilities and current workload before data processing begins. This pre-organization allows data packets to be routed directly to appropriate domains without requiring real-time coordination or decision-making, significantly reducing communication overhead and coordination time during actual processing.
Data Source
AI summary
The embodiments of the disclosure provide a computer system, comprising a plurality of computing devices, each computing device comprising a storage unit and at least one computing unit which are connected in parallel, the plurality of computing devices being divided into a plurality of nodes, and the plurality of nodes forming a network topology; a communication device comprising an autonomous domain building module and a first communication module, the first communication module being configured for connecting the plurality of computing devices in a communication manner; and a task processing device comprising an addressing module and a task processing module. Therefore, the solution of the present disclosure meets the explosive demand of massive data and large amount computing power caused by artificial intelligence and blockchain to a certain extent.