Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Multi-CPU brain-like simulation system

A technology for simulating systems and computing nodes, applied in the computer field, can solve problems such as high power consumption, large size, and low efficiency, and achieve the effects of reducing traversal time, reducing the number of table items, and improving scalability

Active Publication Date: 2020-04-10
GUANGDONG UNIV OF TECH
View PDF3 Cites 5 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Although the existing supercomputer has the ability of brain-like computing, its disadvantages such as high power consumption, high cost, large size and low efficiency make it very difficult to carry out brain-like computing research and realize its application.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Multi-CPU brain-like simulation system
  • Multi-CPU brain-like simulation system
  • Multi-CPU brain-like simulation system

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0041] Such as figure 1 As shown, a kind of multi-CPU brain-inspired simulation system described in this embodiment includes a plurality of brain-inspired simulation system motherboards 1, and the multiple brain-inspired simulation system motherboards 1 are connected in an orderly manner; the brain-inspired simulation system motherboard 1 is composed of six Computing nodes 2 are composed of computing nodes 1, 2, 3, 4, 5, and 6 respectively. Computing nodes 2 and brain-inspired simulation system motherboard 1 communicate through SATA interfaces, such as figure 2 shown.

[0042] Such as image 3 As shown, the computing node 2 includes eight CPUs and a routing system; the routing system is composed of FPGA and CAM, and the CPU and FGPA are connected by an RGMII communication interface.

[0043] Due to the huge number of connections between the nodes of the multi-CPU brain-like simulation system, it is necessary to have symmetry to realize the connection between the circuit boa...

Embodiment 2

[0052] Compared with Embodiment 1, the present embodiment 2, CPU is multi-core CPU, adopts single CPU multi-core analog neuron, specifically as follows:

[0053] Single CPU multi-core simulates neurons, that is, uses multiple threads to work together to simulate neurons. To realize single CPU multi-core analog neuron, two key problems need to be solved: the problem of sending data packets to other threads and the problem of addressing received pulse data packets.

[0054] In order to solve these two problems, a routing thread for processing the routing of pulse data packets between neuron threads is bound on one core of the multi-core CPU, a clock synchronization thread is bound on one core, and all other cores are bound There are threads for simulating neurons, an internal buffer for storing pulse data packets generated by neurons in the thread, and a receiving buffer for storing pulse data packets sent by other threads; and the multi-core CPU is provided for sending outside ...

Embodiment 3

[0059] Combined with the hardware platform of the multi-CPU brain-like simulation system, the single-CPU multi-core analog neuron system is further optimized to realize multi-CPU multi-core analog neuron.

[0060] Compared with Embodiment 2, this embodiment 3 has an external receiving cache for receiving the pulse data packets sent by the routing system, and at the same time increases the network service function based on the original socket for receiving / sending data into the routing system, such as Figure 9 shown. And add an outgoing flag bit in the forwarding flag variable of the neuron, when this bit is 1, it means sending the pulse data packet outside the CPU.

[0061] The forwarding process of the pulse data packet from the thread to the routing thread is the same as that of a single CPU multi-core analog neuron. The difference is that when judging the flag variable, it is necessary to judge whether the outgoing flag is 1, and if it is 1, the pulse data will be forwar...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

The invention discloses a multi-CPU brain-like simulation system which comprises a plurality of brain-like simulation system mainboards which are connected in order. The brain-like simulation system mainboard is composed of six computing nodes, and the computing nodes and the brain-like simulation system mainboard are communicated through an SATA interface. The computing node comprises a pluralityof CPUs and a routing system. The routing system is composed of an FPGA and a CAM. The CPU and the FGPA are connected through an RGMII communication interface. The computing nodes in the same brain-like simulation system mainboard are in logical connection, a regular hexagon interconnection structure is adopted, and each edge of a regular hexagon represents one logical connection. Based on the routing system composed of the FPGA and the CAM and the regular hexagonal interconnection structure, a large number of CPUs are effectively connected, so that the number of physical connections is reduced while normal communication between computing nodes is achieved, the system implementation difficulty is reduced, and good expansion capacity is achieved.

Description

technical field [0001] The invention relates to the technical field of computers, in particular to a multi-CPU brain-like simulation system. Background technique [0002] The human brain has undergone long-term natural development and biological evolution, and gradually formed a super logical thinking ability and excellent intelligence perception ability. The human brain can reason about all the things and creatures that human beings come into contact with. The human brain can easily handle various problems through touch, vision, hearing, logical thinking, reasoning, and decision-making strategies. The above-mentioned capabilities of the human brain are incomparable to modern computers, but they are also the goal of continuous improvement and hard work of modern computer technology. To further develop and study intelligent brain-inspired computing, we should use the biological brain mechanism, especially the human brain mechanism, as a reference, and build a general intell...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Applications(China)
IPC IPC(8): G06N3/06G06N3/063G06F9/50
CPCG06N3/061G06F9/5066G06N3/065Y02D10/00
Inventor 刘怡俊梁君泽叶武剑翁韶伟张子文
Owner GUANGDONG UNIV OF TECH
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com