A configurable and high-speed FPGA configuration circuit based on an SELECTMAP and an implementation method

Abstract

The invention relates to a configurable and high-speed FPGA configuration circuit based on an SELECTMAP and an implementation method. The configuration circuit comprises a high-speed AMBA AHB interface module which is used for sampling signals on an AHB bus and returning corresponding control signals and data signals; the register configuration module comprises a plurality of configuration and state registers and is used for configuring various functions of the main controller and registering numerical values of corresponding states, so that the circuit can be suitable for various different application scenarios; the cache module is used for caching the configuration data input from the AHB system, so that the SELECTMAP interface and the AHB system can perform asynchronous operation, and the flexibility and the stability of the configuration circuit are improved; and the main control module is used for generating data meeting the SELECTMAP time sequence requirement, reading the data inthe cache module, performing shift output, and generating a CCLK signal according to the corresponding data signal, thereby realizing discontinuous transmission of the configuration data, saving a chip pin and improving the reliability of the configuration circuit.

Description

technical field [0001] The invention belongs to the field of computer engineering and is closely related to SoC. In particular, it relates to an FPGA configuration circuit and an implementation method. Background technique [0002] With the development of big data and the improvement of computing power, artificial intelligence has ushered in a new round of explosion in the past two years. After Google AlphaGo won the Go game in 2016, artificial intelligence has attracted high attention in the industry and capital circles, and has become a new outlet. [0003] In the era of artificial intelligence, people's demand for computing power has increased exponentially, and the requirements for computing power have exceeded Moore's Law. With the rapid growth of Internet users and the rapid expansion of data volume, the demand for computing in data centers is also rising rapidly. The computing demands of various applications such as deep learning online prediction, video transcodin...

AI Technical Summary

Problems solved by technology

Therefore, the problem of FPGA configuration time in CPU+FPGA is increasingly restricting the development of new fields such as artificial intelligence, big data and cloud computing.

Images