There have been new advancements in ferroelectric storage technology.
According to reports, Chinese scientists have successfully developed a fatigue-free ferroelectric material and published their findings in the top international academic journal, “Science.”
Traditional storage chips are limited by the read/write cycles of ferroelectric materials, and their stability decreases over time, which has long hindered further research and application of storage chips.
The Chinese research team, based on the two-dimensional sliding ferroelectric mechanism, developed a new type of two-dimensional layered sliding ferroelectric material (3R-MoS2). The storage chips made from this material are expected to overcome the read/write cycle limitation and achieve unlimited read/write cycles.
The core of this research is the use of “interlayer sliding” instead of the “ion movement” in traditional ferroelectric materials. Through AI-assisted cross-scale atomic simulation analysis, the microphysical mechanism of the fatigue resistance of the two-dimensional sliding ferroelectric material was revealed.
The following figure shows the fatigue resistance analysis of the 3R-MoS2 ferroelectric device.
Experiments show that the ferroelectric chip devices made of this material exhibit no decay in ferroelectric polarization even after 4 million cycles of electric field reversal polarization, as measured by the electrical curve.
This technology not only greatly enhances the reliability and durability of storage chips but also helps reduce costs, increase storage density, and has the potential for future applications in extreme environments such as aerospace and deep-sea exploration, as well as in wearable devices and flexible electronic technologies.
Value of Ferroelectric Memory
Ferroelectric memory is a type of random access memory that combines the fast read/write and data retention capabilities of DRAM with the characteristics of other stable storage devices. It is not as dense as DRAM and SRAM, but it has fast storage speeds with low power requirements.
Although it may not replace these technologies, it has broad application prospects in small devices such as PDAs, mobile phones, power meters, smart cards, and security systems. The development of ferroelectric memory began in 1921, and in 1993, the American company Ramtron International successfully launched the first 4K-bit FRAM product. Its technical characteristics differ from floating-gate memory, as it uses artificially synthesized lead zirconate titanate materials and stores data through the polarization charge formed by the reversal of ferroelectric domains under an electric field. It has advantages such as non-volatility, no erase-write delay, fast write speed and unlimited write life. Its principle is to use the ferroelectric effect of ferroelectric crystals to achieve data storage, with features of non-volatility and resistance to magnetic field interference.
Ferroelectric storage technology was proposed as early as 1921, and in 1993, Ramtron International successfully developed the first 4K-bit ferroelectric memory FRAM product, with all FRAM products being manufactured or licensed by Ramtron. FRAM has new developments, adopting a 0.35um process, launching 3V products, and developing single-transistor single-capacitor FRAM cells with a maximum density of up to 256K bits.
FRAM uses the ferroelectric effect of ferroelectric crystals to achieve data storage. The ferroelectric effect refers to the movement of the central atom of the crystal under a certain electric field to reach a stable state. When the electric field is removed, the central atom remains in its original position. This is because the middle layer of the crystal is a high-energy barrier, and the central atom cannot cross the barrier without external energy to reach another stable position. Therefore, FRAM maintains data without the need for voltage or periodic refreshing like DRAM. Because the ferroelectric effect is an intrinsic polarization characteristic of ferroelectric crystals and is not related to electromagnetic effects, the contents of FRAM memory are not affected by external conditions such as magnetic fields, and it can be used like regular ROM memory with non-volatile storage characteristics.
FRAM features fast speed, low read/write power consumption, and no maximum write cycle issues like EEPROM. However, due to the limitations of ferroelectric crystals, FRAM still has a maximum number of read cycles.
FRAM products combine the advantages of RAM and ROM, with fast read/write speeds and non-volatile memory characteristics. Due to the inherent limitations of ferroelectric crystals, the access cycles are limited, and beyond the limit, FRAM will no longer be non-volatile. However, this does not mean that FRAM will be scrapped after exceeding the limit; it just loses its non-volatility but can still be used like ordinary RAM.
| FRAM | EEPROM | Flash memory | SRAM | |
|---|---|---|---|---|
| Memory Type | Non-volatile | Non-volatile | Non-volatile | Volatile |
| Data Backup Battery | No | No | No | Yes |
| Guaranteed Write Cycles | 10 trillion | 1 million | 100 thousand | Unlimited |
| Write Method | Overwrite | Erase + Write | Erase + Write | Overwrite |
| Write Cycle Time | 150ns | 5ms | 10μs | 55ns |
| Booster Circuit | No | Yes | Yes | No |
| Dynamic bit-wise programmable | Yes | No | No | Yes |
| Write speed (13KB) | 10ms | 2 secs | 1 sec | ＜10ms |
| Unified memory (Flexible code and data partitioning) | Yes | No | No | No |
FRAM and EEPROM: FRAM can be a second choice to EEPROM. It has all the performance of EEPROM but with much faster access speeds. However, before deciding to use FRAM, it must be ensured that exceeding the 1 million access cycles in the system will not pose a risk.
FRAM and SRAM: From the perspective of speed, price, and ease of use, SRAM is superior to FRAM. However, in terms of overall design, FRAM has certain advantages. Suppose the design requires about 3K bytes of SRAM and several hundred bytes to store startup code configuration in EEPROM. Non-volatile FRAM can save the startup program and configuration information. If the maximum access speed of all memory in the application is 70ns, a single FRAM can be used to complete this system, making the system structure simpler.
FRAM and DRAM: DRAM is suitable for situations where density and cost are more important than speed. For example, DRAM is the best choice for graphics display memory, where many pixels need to be stored, and recovery time is not very important. If it is not necessary to save the content from the last session after a reboot, volatile DRAM memory can be used. The function and cost of DRAM are incomparable to FRAM, and it has been proven that DRAM cannot be replaced by FRAM.
FRAM and Flash: The most commonly used program memory is Flash, which is very convenient to use and increasingly cheap. Program memory must be non-volatile, relatively inexpensive, and easy to rewrite, but FRAM’s access cycle limitation means that after multiple reads, it will lose its non-volatile characteristics.
Disclaimer: This article is created by the original author. The content of the article represents their personal opinions. Our reposting is for sharing and discussion purposes only and does not imply our endorsement or agreement. If you have any objections, please get in touch with us through the provided channels.
