On-chip super-large-scale power supply network parallel simulation method based on spectrogram rarefaction

A power supply network, ultra-large-scale technology, applied in electrical digital data processing, special data processing applications, instruments, etc., can solve the problem of low matrix parallel efficiency, and achieve the effects of high parallel efficiency, low memory consumption, and fast convergence speed.

Pending Publication Date: 2022-02-01
TSINGHUA UNIV
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] This application provides a parallel simulation method and device for ultra-large-scale power supply network based on spectral graph sparsification, to solve the problem that the existing spectral graph sparse algorithms are implemented serially, although the sparse matrix-vector multiplication and vector Addition is easy to parallelize, but the Cholesky decomposition of sparse and irregular matrices and the parallel efficiency of the previous generation and back generation process are low.

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
  • On-chip super-large-scale power supply network parallel simulation method based on spectrogram rarefaction
  • On-chip super-large-scale power supply network parallel simulation method based on spectrogram rarefaction
  • On-chip super-large-scale power supply network parallel simulation method based on spectrogram rarefaction

Examples

Experimental program
Comparison scheme
Effect test

Embodiment Construction

[0040] Embodiments of the present application are described in detail below, examples of which are shown in the drawings, wherein the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below by referring to the figures are exemplary, and are intended to explain the present application, and should not be construed as limiting the present application.

[0041] figure 1 It is a flow chart of a parallel simulation method for a very large-scale power supply network based on spectral graph sparsification provided according to an embodiment of the present application.

[0042] Such as figure 1 As shown, the parallel simulation method of ultra-large-scale power supply network based on spectral graph sparsification includes the following steps:

[0043] In step S101, the SPICE netlist of the power supply network is read, and the Laplacian matrix and weighted undirected graph cor...

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 super-large-scale power supply network parallel simulation method and device based on spectrogram rarefaction, and the method comprises the steps: reading an SPICE netlist of a power supply network, and building a Laplacian matrix and a weighted undirected graph corresponding to the SPICE netlist, and a right-end item; running a parallel spectrogram sparsification algorithm on the weighted undirected graph to obtain a sparse sub-graph and a Laplacian matrix corresponding to the sparse sub-graph; decomposing the Laplacian matrix corresponding to the sparse sub-graph by using a region decomposition method to obtain an integral Schel complement matrix; and setting a convergence threshold, solving a power supply network linear equation set which takes a Laplacian matrix corresponding to the SPICE netlist as a coefficient matrix based on an integral Schel complement matrix and a convergence threshold operation precondition conjugate gradient method, and obtaining power supply network simulation results of node voltage and the like. Therefore, the problems that an existing spectrogram rarefaction algorithm is achieved in a serial mode, and even though sparse matrix vector multiplication and vector addition in the PCG iteration process are prone to parallel, the parallel efficiency of the Cholesky decomposition process, the previous substitution process and the back substitution process of a sparse and irregular matrix is low are solved.

Description

technical field [0001] The present application relates to the technical field of physical verification and analysis of integrated circuits, and in particular to a method and device for parallel simulation of ultra-large-scale power supply network in a chip based on spectrogram thinning. Background technique [0002] In the process of IC backend design, in order to verify the correctness of the design, it is necessary to simulate the circuit. As a kind of circuit simulation, the simulation of power supply network is an important step in the back-end design process of integrated circuits. The power supply network is the interconnection network that connects the external power supply to all the transistors on the chip. As the process node enters below 7nm, the number of transistors on the chip becomes more and more, the power supply network becomes more and more complex, and the scale of the matrix that needs to be solved in the simulation of the power supply network becomes l...

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): G06F30/367G06F30/327G06F17/16G06F16/901
CPCG06F30/367G06F30/327G06F17/16G06F16/9024
Inventor 喻文健刘志强
Owner TSINGHUA UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products