Flash memory

Flash memory is a type of electronic non-volatile computer memory. Non-volatile means it can hold onto data even when the power is turned off.

This technology is essential for modern life. It provides a way to store information in a solid state without moving parts. Because it lacks mechanical parts, it is much more resistant to physical shock than older hard drives. Today, it is the dominant way to store data in many digital systems.

At the heart of flash memory is a tiny component called a floating-gate MOSFET.

This component acts as a memory cell. To store information, electrons are moved into a floating gate. This process can be done through hot-electron injection. To erase the data, a process called quantum tunneling is used to move the electrons out.

Some newer versions use charge trap flash (CTF) technology. Instead of a polysilicon floating gate, CTF uses an insulating silicon nitride layer to trap electrons. This can help prevent electron leakage and allow for smaller cells.

There are two primary types of flash memory: NOR flash and NAND flash. They are named after the logic gates used in their circuit designs.

NOR flash allows for direct random access. This means a computer can read a single byte of data very quickly from any location. This makes NOR flash perfect for executing code, such as a computer's BIOS. NAND flash, however, uses a serial access approach. It reads and writes data in large blocks or pages rather than individual bytes.

This makes NAND flash much better for high-density, mass storage.

The history of this technology began with the invention of the MOSFET at Bell Labs around 1959. In 1967, Dawon Kahng and Simon Min Sze developed the floating-gate MOSFET. This paved the way for early memory like EPROM and EEPROM. However, early EEPROM was slow and expensive because engineers had to build a cell for every bit. In 1980, Fujio Masuoka invented flash memory at Toshiba. His colleague, Shōji Ariizumi, suggested the name "flash" because the erasure process reminded him of a camera flash. Toshiba began marketing NAND flash in 1987, and Intel introduced the first commercial NOR flash chip in 1988.

Flash memory is used in a massive variety of devices. You can find it in smartphones, digital cameras, and USB drives. It is also used in medical electronics, industrial robotics, and video game consoles. Because NAND flash is cost-effective and can hold a lot of data, it is the standard for solid-state drives (SSDs). Even small devices like microSD cards use it. A microSD card is incredibly tiny, measuring just over 1.5 cm² with a thickness of less than 1 mm.

While flash memory is powerful, it has limits regarding endurance. Every time a block of memory is erased and rewritten, it undergoes wear. This wear is caused by the degradation of the oxide layers that insulate the cells. If these oxides degrade, they may lose their ability to prevent electron leakage. This can lead to data loss. Depending on the type, a chip might endure anywhere from 100 to 1,000,000 erase cycles. This is why managing how often data is written is important for the life of the device.

To increase storage capacity, engineers use advanced stacking techniques. One method is 3D integrated circuit (3D IC) technology, which stacks chips vertically.

Another method is 3D V-NAND, which stacks the memory cells themselves vertically within a single chip. By using die stacking with through-silicon vias, manufacturers can create very high capacities. For example, a single package can hold up to 1 tebibyte of data by using 16 stacked dies. This allows us to fit massive amounts of information into tiny, portable spaces.