FPGA (field programmable gate array) is the product of further development based on pal, gal, CPLD and other programmable devices. As a kind of semi custom circuit in the field of application specific integrated circuit (ASIC), it not only solves the shortcomings of custom circuit, but also overcomes the shortcomings of limited gate circuits of original programmable devices.
With the development of artificial intelligence and embedded industry, many industry terms are becoming more and more familiar. This is what we call FPGA today. But what is FPGA? What are its advantages? Many people are not so clear.
What is FPGA?
FPGA is the abbreviation of field programmable gate array. It is a kind of cheap device that developers can use personal PC to realize their desired functions in a short time, and it can be rewritten many times.
In fact, FPGA has been supported by most users all over the world. In 2005, according to the survey report provided by iSuppli, two of the world’s top five sales of ASIC and PLD (including FPGA programmable products) were FPGA suppliers. In the industry, Xilinx of the United States ranks first in the world, IBM ranks second in the world, and Altera of the United States ranks fourth in the world.
Xilinx pioneered the innovative technology of FPGA, and launched the commercial product “xc2000” for the first time in 1985.
At that time, CMOS LSI was the mainstream of semiconductor devices, and large-scale logic systems appeared one after another“ “Gate array” was the simplest way to LSI the logic system at that time. Gate array is a kind of semi custom chip, which is prepared by semiconductor manufacturers to form a logic gate called basic unit on the silicon chip. By wiring according to the user’s desired circuit, the circuit is formed on the motherboard. Although the cost and development cycle of ASIC is the shortest, it still needs several weeks of development time.
On the other hand, the field programmable device PLA derived from prom is widely used. PLA is a logic device, which can be programmed by the user. PLA is a logic LSI characterized by integrated structure. It can write connection information through the same device as Prom program. Then, new devices such as EPROM and EEPROM appeared one after another, which were popular because they were practical and easy to use. However, it is still difficult to implement arbitrary large-scale logic freely, and they can only play an auxiliary role.
At the beginning of the birth of FPGA, computer researchers were the first to feel the great advantages of FPGA. Before that, in order to realize the design idea of a new computer architecture, we must develop an ASIC as a prototype and install many bulky ICs on the panel. However, unlike these methods that require huge cost and effort, once users create a prototype board with multiple FPGAs, they can immediately run your new architecture. And it’s easy to fix and change specifications. Therefore, many new architectures emerge as the times require. Therefore, the research on reconfigurable processor and new FPGA architecture also begin to flourish. After that, the characteristics of FPGA’s flexible use and strong applicability are also highly praised in the field of communication / image processing, and are applied to various devices of communication network such as router.
Programming technology of FPGA
At the beginning of 2000, SRAM was the mainstream FPGA programming technology at that time, in addition to other types of FPGA programming technology.
In the late 1980s, a technology with a different concept from SRAM FPGA appeared, which is called “fuse technology” (anti fuse technology). In the fuse technology, the connection point of the circuit is the fuse, and the circuit is formed by burning the fuse. Antifuse element is a memory element that can only be programmed once. Once programmed, it will conduct with low impedance. It is said that it is more compact than SRAM devices and can realize high-speed operation with high integration and low impedance. The disadvantage of fuse technology (anti fuse technology) is that it cannot be rewritten and used repeatedly. In addition, it is difficult to reduce the price is also its disadvantage.
At the same time, another new device “CPLD” was released. This is a PLD that can realize any large-scale logic circuit by integrating multiple small PLDs into one LSI. Like small-scale PLD, nonvolatile and rewritable ease of use are the biggest characteristics of CPLD. However, the design freedom of CPLD is not as high as FPGA.
In the 1990s, “flash type FPGA” using flash memory appeared. Flash FPGA has two advantages of SRAM technology and antifuse technology, namely rewriting and nonvolatile. However, due to its complex structure and manufacturing process, it is difficult to improve the integration and reduce the price. Therefore, flash is not used in large-scale and high-speed FPGA.
Architecture of FPGA and PLD
FPGA and CPLD based on SRAM are the mainstream technology at present. FPGA based on SRAM is the most suitable for high-speed and large-scale applications. Either way, it has the advantages of simple manufacture and low price.
FPGA design is not a simple chip research, mainly using FPGA mode to design products of other industries. Different from ASIC, FPGA is widely used in communication industry. Through the analysis of the global FPGA product market and related suppliers, combined with the actual situation of our country and domestic leading FPGA products, we can find the development direction of related technologies in the future, which has a very important role in promoting the comprehensive improvement of China’s scientific and technological level.