What is Flash Memory?

Flash memory differs from previous memory because only one cell can be written to or erased at a time.

However, advanced EEPROM can process multiple cells simultaneously using electrical impulses.

What is Flash Memory for Data Storage?

Therefore, systems can read and write to different parts of the flash memory simultaneously. This allows flash to operate at much higher speeds.

From an economic perspective, 4, 8, and 16 GB devices on the market sell for around 50 TL. However, if you want higher capacity and speed, the price increases almost tenfold.

In contrast, the cost per GB of hard drives is much lower than flash memory. Hard drives also offer significantly higher capacities than flash memory.

Flash memory is shock resistant and consumes low power because it has no mechanical actuators or moving parts, making it silent.

Its small size makes it a preferred choice for portable devices. Its light weight and versatility make it advantageous for many applications.

However, each type of flash memory allows a limited number of write and erase operations. This number depends on the cell, the manufacturing process, and the required erase voltage; it is typically between 10,000 and 1,000,000.

Flash memory is constructed with NOR and NAND logic gates to store 0s and 1s. File systems for these memory devices, such as JFFS for NOR, are used. For NAND, the second version evolved into JFFS2, NAND, or YAFFS.

In practice, the FAT file system is preferred for compatibility with removable memory cards.

Another feature of some memory cards for high-end digital cameras is their high thermal resistance. The supported temperature range is -25°C to 85°C, allowing them to operate in extreme conditions such as deserts or glaciers.

Flash Memory History

The history of flash memory is linked to routers, modems, and PC BIOS. The advancement of wireless technologies has also influenced flash memory.

In 1984, Fujio Masuoka invented flash memory as an evolution of the existing EEPROM. Intel was the first to commercialize flash memory and claim success.

Between 1994 and 1998, mainstream memory types such as SmartMedia and CompactFlash were developed. This led to applications in other areas.

In 1998, Rio commercialized the first Walkman without moving parts. This device utilized the SmartMedia operating mode.

In 1994, SanDisk began marketing memory cards (CompactFlash) based on these circuits. Since then, the evolution has extended to small portable devices such as portable MP3 players, video consoles, and PC Cards.

How Does a USB Memory Work?

The EEPROM type of flash contains an advanced transistor cell matrix. These cells store only one bit of information.

In addition, multi-level cell devices have been developed, allowing multiple bits to be stored per cell.

These memories are based on the FAMOS transistor. Essentially, they are NMOS transistors with an additional conductor. Furthermore, electrons are stored between the control gate (CG) and the floating gate (FG).

NOR Type

In NOR flash memories, if the CG is activated while the electrons are in the FG, the electric field changes. Thus, the cell exhibits an electric field based on whether it is in a 1 or 0 state. Thus, when a voltage is applied to the CG, current flows or not according to the voltage stored in the cell.

The presence or absence of current is detected and interpreted as a 1 or 0. In multi-level cells, the number of electrons in the FG is controlled and interpreted by the current density.

To program a NOR cell, current is passed from the source terminal to the sink terminal. Then, a high voltage is applied to the CG, and electrons are captured. This process is called hot electron injection.

The Fowler-Nordheim tunneling technique is used to erase a cell. It turns the transistor into an electron gun, and electrons are drawn by applying a reverse voltage. Such a high voltage can cause cell damage in thin conductors.

Flash memories are divided into blocks. Since the erasing process is slow, all blocks are cleared, speeding up the process. Therefore, flash memories are much faster than traditional EEPROMs, which are erased byte-by-byte. However, the block must be cleared before the data can be rewritten.

NAND Type

Flash memory based on NAND logic gates operates differently. This is because an injection tunnel is used for writing and a drop tunnel is used for erasing.

NAND-based memory is ten times more resilient than other gates. However, it only allows sequential access compared to NOR-based flash memory.

However, the erasing mechanism is simpler, providing a cost-effective basis for memory card-type devices.

Popular USB sticks, pen drives, and SSDs use NAND-type flash drives.

Comparison of NOR and NAND-Based Memory

To compare these types of memory, different aspects of traditionally valuable memories are taken into account.

1. The storage density of the chips is currently relatively high in NAND memory.
2. NOR cost is much higher.
3. NOR access is random for reading and block-oriented for modification. However, NAND only offers direct access to blocks and sequential reading within blocks.
4. When writing NOR, we can change a single bit. This is highlighted by the limited reprogramming of NANDs that must complete blocks or words.
5. Reading speed is much higher in NOR than in NAND.
6. The write speed for NOR is five µs per byte and 200 µs per page in NAND.
7. The deletion rate for NOR is 1s per 64 KB block and 2ms per 16 KB block in NAND.
8. NOR-based devices are very reliable. They are also resistant to data corruption. Unlike NAND systems, no false blocks are created. Therefore, NOR systems are more reliable.

In short, NAND systems are cheaper and faster, but they lack reliability and are essential for a good file system. Depending on what you’re looking for, it makes sense to choose one over the other.

File Systems

Designing an efficient file system for flash memory is complex because NOR and NAND flash memory types have different characteristics.

In addition, a file system designed for NOR includes unnecessary mechanisms for NAND. Furthermore, managing NAND requires additional mechanisms.

An example is the garbage collector. Erasing in NOR is slow and complex. Furthermore, this limits file system options. However, NAND systems erase much faster, so these limitations are meaningless.

Another difference is that NAND can have bad blocks. In NOR systems, integrity is guaranteed, and bad blocks are meaningless.

The size each system must handle varies. Therefore, the design must be tailored to the functions you intend to provide the system.

Background

Memory has evolved significantly since the dawn of computing. The types of semiconductor memory used as main memory are also important. However, it’s helpful to remember a few points about flash memory and its usage context.

   Read-Only Memories

• ROM (Read Only Memory): These are mainly used for microprogramming systems. Manufacturers often use them when using mass-produced ingredients.
• PROM (Programmable Read-Only Memory): The writing process is electronic. Unlike the previous ones recorded during production, the chip can be saved after it has been produced. It allows for a single recording and is more expensive than ROM.

   All Read Memory

• EPROM (Erasable Programmable Read-Only Memory): It can be written electronically several times, but the deletion of the content is wholly and fully exposed to ultraviolet rays.
• EEPROM (Electrically Erasable Programmable Read-Only Memory): You can delete a byte by selecting it with an electric current. It is more expensive than the EPROM.
• Flash Memory: It is based on EEPROM memories but allows block deletion, is cheaper, and is more intense.

   Read/Write Memories

• DRAM (Dynamic Random Access Memory): The data is stored on the charge of a capacitor. It tends to discharge, and therefore, a periodic renewal process is required. They are simpler and cheaper than SRAMs.
• SRAM (Static Random Access Memory): The data is stored on flip-flops, so a refresh is not required. Like DRAM, this is volatile. They are faster and more expensive than DRAMs.

The Future

Smart and integrated devices are becoming ubiquitous. Therefore, the future of flash memory looks promising. Furthermore, cheap and flexible memory will meet both the feature and cost requirements of new systems.

It appears that the miniaturization and density of flash memory have not yet reached physically alarming levels. However, with the advent of the memristor, the future of flash memory is beginning to take shape.

USB memory development is very rapid. It surpasses other memory types in capacity, speed, and performance. However, communication standards, especially in PCs, are still low and may hinder progress.

Companies such as AMD and Fujitsu are investing heavily in research and development. These investments supported the establishment of new companies dedicated to flash memory between 2003 and 2005. Furthermore, combining two technologies in the flash drive world is a technical achievement.

Flash drives continue to specialize by utilizing each type of memory for specific functions.

For example, in PDAs, Harvard Architecture and ORNAND-type memory provide high read speeds with minimal power. Data memory, on the other hand, offers high capacity with NAND gates.

Critical devices utilize NOR or ORNAND-based technologies. Personal consumer electronics, on the other hand, continue to favor NAND memory.

This is because it is low-cost and suitable for devices such as portable MP3 players and DVD players.

Using lower voltage reduces power consumption and extends device life. However, miniaturization and higher processor clock frequencies create new challenges.

Wireless integration is enabling the proliferation of digital devices and transforming the world around us.

Security Steps to Consider

First, disable the autorun.inf feature on USB devices. Also, ensure you have system administrator permissions. You need to modify the nodrivetypeautorun value.

You can do this using regedit.exe. Launch it from the Start menu or Run. Then, locate and open the nodrivetypeautorun value.

Select the replace option and replace the default value of 91 with the hexadecimal 95. This will limit the execution of malicious code on external devices.

However, this method does not prevent execution when accessing the content indirectly through explorer.exe. In this case, use the folders option within the MIPC and select the drive.

Analyze the contents of the root directory. Also check for the presence or modification of autorun.inf. Malware creators may hide this file.

Taking appropriate precautions will make it easier to prevent the spread of malicious programs next time. However, this method is only effective if the computer is not infected.

Otherwise, the malicious code will be recreated in memory each time it is deleted. Remember, infected devices can infect other computers. Re-infection is possible.