Nor NAND and onenand flash difference Summary

[What is onenand flash]

Onenand is a highly reliable embedded storage device designed for consumer electronics and the next-generation mobile phone market.

With the development of NAND technology over the past few decades, some companies, based on the original NAND architecture,

An ideal single storage chip is designed, which integrates the SRAM cache and logic interface.

Onenand not only achieves fast reading speed of nor flash, but also retains the advantages of NAND Flash's large-capacity data storage.

What corresponds to onenand is the NAND Flash and nor flash that appeared earlier.

[Features of onenand Flash]

Compared with the other two types, it is easier to see the features of onenand:

Differences between onenand NAND nor flash

Application Requirements	Nand	Onenand	Nor
Fast random read			√
Quick sequential read	√	√	√
Fast writing/programming	√	√
Multiple blocks are erased simultaneously.		√ (Up to 64 blocks)	√
Erased suspension/restoration		√	√
Write back	√ (Error detection)	√ (Error detection and correction)
Lock/unlock/lock		√	√
Error Correction	External (hardware/software)	Built-in	No
Scalability	√	√

NAND Flash and nor Flash are two major non-easy-to-lose flash chips on the market. Compared with nor flash, the advantages of NAND flash in capacity, power consumption, and service life make it an ideal solution for high data storage density. Nor Flash has a high transmission efficiency, but the write and erase speeds are low; with features such as large capacity, fast writing speed, small chip area, high unit density, fast erasure speed, and low cost, NAND Flash boasts strong market competitiveness in non-volatile storage devices.

Structure: nor flash is parallel, and NAND Flash is serial.

Bus: nor Flash is a separate address line and data line, while nandflash is reusable.

Size: the typical NAND Flash size is 1/8 of the nor flash size.

Bad blocks: Bad blocks in the NAND device are randomly distributed and need to be initialized and scanned to find bad blocks and marked as unavailable.

Bit switching: NAND flash occurs more frequently than nor flash. We recommend that you use the EDC/ECC algorithm when using NAND Flash.

Usage: nor flash can be executed in the chip (xip, execute in place). Applications can run directly in fIash flash memory without having to read the code into system Ram; NAND Flash requires an I/O interface, so the driver needs to be written for use.

Through the above analysis and comparison, NAND Flash is more suitable for Embedded Systems with large data storage capacity.

In around 1984, during his employment at Toshiba, Dr. fujiio Masuoka invented a unique storage device with read-only memory (ROM), random access memory (RAM) ideal features of the read-only memory (EEPROM) device. Like Ram and EEPROM, you can write, erase, and rewrite data on these novel devices or similar Rom devices. The devices can save recoverable static data for almost unlimited time. In addition, compared with Ram and Rom, Dr. Masuoka's memory capacity is easy to increase. It is important that, compared with ROM and EEPROM, the chip can store data for a long period without backing up the battery or external power (see figure ).

In addition to the absence of external power (non-volatile) to store data and the ability to write and erase information, the emergence of new types of memory with faster read/write operations and lower costs, EEPROM and Rom are no longer so common. Shoji Ariizumi, a colleague of Dr. Masuoka, compares the erasure speed of the new memory to the Flash speed of the camera. Although it is not as fast as Ram, it is quite attractive and named after the flash memory.

Later that year, flash was unveiled at the IEEE 1984 International Electronic Devices conference in San Jose, California. After predicting the benefits and economic vitality of the technology, Intel launched its first commercial nor flash chip in 1988. Nowadays, flash memory is widely used in large-capacity consumer electronic products, such as digital audio players, cameras, mobile phones, USB drives, memory cards and video games, and complex design such as embedded systems and integration in microcontroller.

Classification of flash memory

Flash Memory consists of two types: nor and NAND. At work, nor memory is used as a typical Ram operation in a computer, allowing direct access to a single byte or multi-byte space regardless of their location in the bucket. In terms of usage, nor flash memory can be used for specialized software for storing components, such as router firmware or computer BIOS. Nor flash memory requires a life cycle of about 100,000 write times. If the life cycle is exceeded, a bad zone may occur.

About a year after intel exhibited its first nor flash device, Toshiba developed NAND Flash, which relies on flash conversion software to make the device a hard drive similar to an operating system. Compared with nor, this flash memory has three obvious advantages: long life, 1 million read/write cycles; faster read/write operations; lower cost. In addition, nand flash memory can retain larger data blocks for long or short-time storage. Nand devices are more useful for storing data collected or downloaded from products, such as data recorder information, photos/videos of digital cameras, and music files of MP3 players.

The third type of flash memory to be updated is the onenand flash memory, developed by Samsung, which supports faster data throughput and higher density, these two points meet the two main requirements of high-resolution photography, video and other media applications. Onenand can be seen as a mixture of nor and NAND technologies. Essentially, a separate onenand chip integrates a nor flash interface, a NAND Flash Controller logic, a NAND Flash array, and a buffer ram of up to 5 kb. As for the speed, it can transmit data at a continuous read rate of up to 108 Mb/s.

The onenand device has two types: muxed and demuxed. For the muxed type, the address pin and data pin are combined, while the demuxed type chip pins are separated. When you focus on reducing the number of pins, choosing muxed onenand may be better. In addition, muxed onenand only works at 1.8 V, demuxed has a low density, less than 1 Gbit, demuxed has two options: 1.8 V and 3.3v. If the density of the muxed or demuxed device exceeds 1 Gbit, you can only select the operating voltage of 1. 8 V.

In short, nor flash memory is suitable for code storage, that is, firmware, device applications, and so on, while nand flash memory is similar to the routine work of hard drive with large storage capacity. Both the onenand flash memory have the advantages of being competent for code and massive data storage, while at the same time being more efficient.

Basic flash work

All types of flash memory contain a large number of units, each of which is a set of floating gate transistor arrays. Early flash devices store one-bit information per unit. However, the continuous development of multi-level single-component, by using more than two levels of charge, the storage capacity of these multi-level single-component exceeds 1 per unit.

The typical nor unit looks like two gates of a MOSFET: A control gate and a floating gate, with a layer of oxide separating the two. The floating gate is located between the substrate and the control gate, where data is stored. When the current flows from the source pole to the drain pole, a high voltage is generated on the control gate, and the nor unit is programmed. When the high voltage reaches an appropriate level, the electronic (data) flows to the floating gate, which is stored by means of insulation layer data. This process is called Hot electronic injection.

With a process called tunnel release, you need to add a large electrical pressure difference between the source pole and the control gate to erase the unit and reset data for all units. This voltage difference is obtained through an integrated charge pump that forces the Electronic Release of the floating gate to erase the unit.

The operating voltage of the nand flash memory chip is 3.3 V or 5 V, and the data is erased using the same tunnel release process as the nor device. When writing data, tunnel injection, that is, the quantum tunnel effect, which is known as the Fowler-nordheim tunnel effect, is a method to inject the Charge Carrier into the floating gate through the oxide insulator. It is said that this method can save power and shorten the write operation time. The comparison of the same performance parameters of NAND and onenand flash memory is shown in the table.

Memory limitations

The most critical limitation of flash memory is the limited number of write/erase cycles. Most commercial flash-based products ensure up to 1 million Write cycles. This number seems very big, probably for nor flash, because it may be okay to save a software or BIOS for a long time. However, in typical NAND applications that frequently write, retrieve, and rewrite files, these cycles are very short, and most users may not count. Flash memory may not be suitable for storing frequently updated key data.

To address this limitation, you can use a firmware or a file system drive to calculate the number of times the memory is written. These software will dynamically remap these blocks and share write operations in the sector. In other words, if the write operation fails, the software authorizes the write operation to unused sectors through write verification and re- ing.

Like Ram, flash memory can read or program one byte or one word at a time, but the erasure must be a complete block at a time, and all the bits in the block must be reset to 1. This means that you need to spend more time programming. For example, if you want to write a bit (0) into a block, you must completely erase the block instead of simply overwriting it.

Advantages of flash memory

The advantages of flash memory far exceed its limitations, which can be proved by the flash-based products on the market and their popularity. Like traditional storage and storage devices, the size of flash memory is initially small and is now close to the level of hard drive in the laptop. For example, about last year, Samsung launched a 32 GB flash drive based on its 32 MB flash memory chip (k9f5608u0bycb0) using NAND technology. Not long ago, A 32 GB chip developed using a 40nm process is introduced. Bitmicro's edisk is another example of a hard drive terminal that provides 3.5 inch GB of storage capacity on 155 solid state disks.

Memory cards, a popular storage medium for digital cameras and other portable products, can now be equivalent to what is considered to be the latest hard drive in the end of 1990s. At present, 1 GB and 2 GB capacity have replaced MB and lower memory cards. The same is true for USB flash drives, also known as thumb drives and/or pocket drives.

In general, flash memory has five obvious advantages: (1) small size, which may be further reduced over time; (2) low power; (3) high storage capacity, over time, it will increase exponentially; (4) it will hardly be destroyed (no moving parts) If encapsulated properly; (5) it will become cheaper.

Flash Road Map

Obviously, Flash is one of the technologies that will accompany us for a while in the future. The new storage technology evolved from this technology also came to light.

Just three months ago, the mmca (multimediacard Association) and jedec solid state Technology Association agreed to choose emmc, as a trademark and product type for Embedded Memory Module products for Embedded flash applications, built on the common mmca/jedec mmc specification. The emmc label details the architecture consisting of an MMC interface, flash memory, and a controller consisting of an embedded storage system, all of which are encapsulated in a small BGA Package. This architecture is expected to win the favor of many products, including industrial products, mobile phones, navigation systems, media players and other portable electronic devices.

Based on MMC system specifications v4.1/4.2 and jedec bga, the system specifies interfaces with a speed of up to 52 Mb/s. It is important that the standard supports an interface voltage of 1.8 V or 3.3v to overcome the current limitations of some flash memory operating voltage.

With emmc, the host system can access all large-capacity memory, including memory cards and hard drives, through an MMC Interface Protocol bus. The system architecture is far more flexible than the architecture based on other memory card standards. Therefore, the standardized emmc protocol interfaces maintain complexity, such as the functional differences of nand flash memory, which are invisible to hosts. In addition, because emmc is an industrial standard, there are multiple sources of memory elements.

In addition, Qualcomm has announced that it uses Samsung's onenand flash memory for all the forthcoming mobile base station modem chipset. There have been a variety of chipsets that support the fastest read/write memory, and more emerging multimedia products will be widely supported.

The main target of the onenand-based chipset is 3G phone. The memory with a write speed of 17 Mb/s ensures wireless download of continuous data streams that exceed the HSDPA specification. In addition to the multimedia mobile phone design, onenand must be a valuable choice for hybrid hard drives with non-volatile buffers. At present, the 60 nm technology chip can provide up to 2 GB of storage capacity, and later this year, it may appear that the use of nm technology capacity of 4 GB chips.

From http://bbs.ednchina.com/BLOG_ARTICLE_1989648.HTM