A multi-channel high-bandwidth data exchange method based on fpga stacking

Abstract

The present invention provides a multi-channel high-bandwidth data exchange method based on FPGA stacking. The equipment used in the method includes switching boards, M input boards, K output boards and synchronous control boards. The method specifically includes the following steps : 1) each data encapsulation unit in the M input boards is processed according to the data encapsulation protocol to the data stream encapsulation and packaging of each data input signal; 2) the data slice and the distribution unit are used for data Each data packet sent by the encapsulation unit is divided into N parts and transmitted to N FPGA data slice exchange units respectively; ...; 5) The data sending unit sends the data slice recombination unit to The incoming data packet removes the encapsulation information including the header and restores the original data signal output. The data exchange method of the present invention breaks through the bottleneck of the total port scale limited by the single-chip port scale, has a simple and reasonable structure, and greatly improves single-port bandwidth and transmission efficiency.

Description

technical field [0001] The invention belongs to the technical field of video processing, in particular to a multi-channel high-bandwidth data exchange method based on FPGA stacking. Background technique [0002] With the continuous advancement of technology, especially in the field of professional audio-visual, with the popularization and explosive growth of high-definition, ultra-high-definition and very high-definition video and other large-scale data services, the market has proposed more bandwidth and scale of high-speed data switching matrix. high demand. The traditional switching matrix has been greatly challenged in terms of scale and bandwidth. [0003] Traditional data exchange methods are roughly divided into the following three categories: 1) Based on dedicated ASIC exchange chips, such as the Crosspoint series chips of Mindspeed (later acquired by MACOM). The switching scale can be from 16×16 to 288×288, and the single-channel bandwidth ranges from 3.2Gbps to 1...

AI Technical Summary

Problems solved by technology

But all there is weak point in above prior art, main defect has: 1) based on the scheme of special-purpose ASIC switching chip

[0004] In the prior art, there are data exchange systems based on FPGA architecture, such as the invention of "CN201320825955 Digital Video Exchange System". The disclosed technical solution is based on a single FPGA architecture, and the total capacity and single-port bandwidth are limited by FPGA technology and It is difficult to meet the needs of ultra-large-scale and ultra-high single-port bandwidth at the same time, and is mainly used for small-scale data exchange

Although it is also mentioned in this document that multiple digital video exchange systems can be formed into a larger-scale data exchange system, this document does not disclose the specific structure of this multiple digital video exchange systems with a single FPGA architecture, and this This method of multi-system superposition must also have the problems of complex structure, duplication, low transmission efficiency and fault tolerance, and no specific data exchange method is provided.

Images