
Click the “blue text” above to follow for more exciting content.
This article contains 496 words, and it takes about 2 minutes to read.
FPGA is mostly based on SRAM technology, which often allows it to keep up with the latest technology nodes, ultimately standing out in PLD (SRAM consists of a transmission transistor and two CMOS inverters. In SRAM, CMOS transistors are used as the basic building blocks of storage cells. By controlling the switching state of the CMOS transistors, read and write operations on the data in the storage cells can be performed). SPLD is mostly implemented using fuse (PROM) and EPROM, EEPROM technology for programmability, while CPLD is mostly based on EEPROM or Flash, which often lags behind the latest nodes by one or more generations. According to the different configuration units, FPGAs can be divided into three categories:
1) FPGA based on SRAM technology, which is the mainstream technology of FPGA. Although the volatility of SRAM requires external storage to retain configuration files, since SRAM is always at the forefront of each generation of CMOS processes, SRAM-based FPGAs can often keep up with the latest technology nodes, a characteristic that sufficiently offsets the need for external storage;
2) FPGA based on anti-fuse technology, which is characterized by the difficulty of transistors flipping when hit by radiation (when a single high-energy particle (such as a neutron, proton, or heavy ion) strikes a transistor (especially its storage cell), it may cause the stored data bits to change erroneously). However, anti-fuse resources are limited and are not suitable for complex signal processing, thus they are often used in aerospace electronic systems to monitor and reconfigure SRAM-based FPGAs. Actel (now a subsidiary of Microchip) is known for its anti-fuse FPGAs;
3) FPGA based on EEPROM or FLASH, characterized by instant-on capability, as they are non-volatile storage that can maintain circuit states once configured, but the technology nodes also lag behind SRAM by one or more generations.
(*My personal understanding of why SRAM technology can keep up with the latest technology: Semiconductor processes are used to describe the width of the transistor gate ➡ the gate is one of the key electrodes in FET or MOS ➡ SRAM consists of a transmission transistor and two CMOS inverters ➡ FPGA based on SRAM technology can often keep up with the latest technology nodes)

Appendix:
|
Term |
Definition |
Main Features |
|
RAM (Random Access Memory) |
It is a component used in computers to temporarily store data, equivalent to the memory stick in a computer. It can be read and written at any time and is very fast, usually serving as a temporary data storage medium for the operating system or other running programs. |
Volatile: RAM is volatile memory, and data in RAM will be lost when the power is turned off. Fast speed: RAM access speed is very fast, almost the fastest among all access devices. High cost: RAM has relatively limited capacity and is relatively expensive. Sensitivity: RAM is very sensitive to environmental static charges, which may interfere with the charge in the capacitors within the memory, leading to data loss or even circuit damage. |
|
DRAM (Dynamic Random Access Memory) |
It is the most commonly used type of memory in computers. It uses transistors and capacitors to store data, but since the charge in the capacitors gradually leaks, DRAM needs to be refreshed periodically to maintain data stability. |
DRAM can have a large storage capacity, making it suitable as the main memory of computers. DRAM access speed is relatively slow because it takes extra time to refresh the storage cells. DRAM has relatively low power consumption because it can reduce power voltage when not accessed to save power. DRAM is relatively low-cost, suitable for mass production and use. |
|
SRAM (Static Random Access Memory) |
It is a type of RAM that uses flip-flops (composed of a transmission transistor (Pass-Transistor, PT) and two CMOS inverters) to store data. SRAM cells utilize the bistability of flip-flops (0 and 1) to record data, controlled by word lines to turn PT on/off. When on, the cell outputs data; when off, the cell outputs the original value. As long as the power remains on, SRAM can continuously hold the stored data without needing refresh operations. |
SRAM access speed is very fast because it does not need to refresh like DRAM. SRAM has relatively high power consumption because it needs to continuously provide power to the storage cells. SRAM has relatively low integration, meaning fewer storage cells can fit on the same size chip. SRAM is more expensive, but due to its high-speed access characteristics, it is often used in situations requiring high-speed caching, such as CPU internal caches. |
|
MOS (Metal-Oxide-Semiconductor) |
It is a semiconductor device based on a metal-oxide-semiconductor structure, also known as a Metal-Oxide-Semiconductor Field-Effect Transistor (MOSFET). In this structure, metal serves as the gate (G, primarily controlling the current flow between the source and drain), the oxide serves as the insulating layer, and the semiconductor forms the source (S, primarily providing current or signal to the FET) and drain (D, primarily receiving and outputting the current or signal amplified or processed by the FET). Classification:MOS can be divided into PMOS (P-channel) and NMOS (N-channel). |
Principle:When a voltage is applied to the metal gate, an electric field is formed in the oxide layer. This electric field can affect the charge distribution in the semiconductor material located beneath the oxide layer. As the voltage increases, the strength of the electric field also increases, causing changes in the charge distribution in the semiconductor, forming a conductive channel. The conductive properties of this channel can be controlled, thereby controlling the current in the MOS. Features:Low power consumption: MOS consumes almost no power in static state, only consuming power during switching state transitions, thus having low power consumption. High speed: MOS has very fast switching speeds, meeting the demands of high-speed electronic devices. High reliability: MOS has good anti-interference ability and stability, capable of functioning normally in harsh environments. Small size: MOS has high integration, small size, making it easy to layout and install in electronic devices. |
|
FET (Field-Effect Transistor) |
It is a transistor that uses the field effect to control the flow of current in semiconductor materials, conducting electricity through the flow of only one type of carrier (either electrons or holes), thus also known as a unipolar transistor. |
The working principle of FET is based on the field effect. Taking the N-channel enhancement-mode MOSFET as an example, when a positive voltage is applied to the gate (G) relative to the source (S), an N-type conductive channel is formed in the P-type semiconductor substrate beneath the gate. This channel connects the source and drain (D), allowing current to flow from the source to the drain. The magnitude of the gate voltage determines the width of the channel and the quality of its conductivity, thereby controlling the current magnitude in the FET. |
|
CMOS (Complementary Metal-Oxide-Semiconductor) |
It is an integrated circuit technology based on field-effect transistors (FET), with its core structure consisting of a pair of complementary transistors, namely N-type MOS transistors (NMOS) and P-type MOS transistors (PMOS). When NMOS and PMOS transistors work simultaneously in a circuit, they can achieve low-power, high-speed logic functions. CMOS chips maintain data by storing charge. When the power is turned off, the charge in the CMOS chip does not disappear immediately, allowing data to be retained for a period. However, over time and with external environmental influences (such as temperature, humidity, etc.), the data in the CMOS chip may gradually be lost, which is why sometimes it is necessary to reset CMOS settings. |
CMOS chips are mainly used to store the basic startup information of computers, such as date, time, startup settings, etc. This information can still be retained after the computer is powered off, until it is read by the BIOS during the next boot to initialize the system. Low power consumption: CMOS circuits consume almost no power in static state, only consuming power during switching state transitions, thus having low power consumption. High integration: CMOS technology can achieve high-density integrated circuit designs, making chips smaller and more powerful. High reliability: CMOS circuits have good anti-interference ability and stability, capable of functioning normally in harsh environments. |

Risk Warning:This content only represents the analysis, speculation, and judgment of the research conducted by the author, and is published here solely for the purpose of conveying information, not as a basis for specific investment targets. Investment carries risks, and one should be cautious when entering the market!Copyright Statement:This content is copyrighted by the original party or author. If reproduced, please indicate the source and author, retain the original title and the integrity of the article content, and bear legal responsibilities such as copyright.
END
