Information Technology

May 25, 2006
Flash memory: Preferred storage media

Geetanjali Babbar

NEW DELHI -- Flash memory allows stored data to be saved even when the memory device is disconnected from its power source. Flash memory stores information on a silicon chip making power unnecessary to maintain the information in the chip. Because of this advantage, flash memory technology has evolved into the preferred storage media for a variety of consumer and industrial devices.

Toshiba invented Flash memory in the 1980s as a new memory technology that allowed stored data to be saved even when the memory device was disconnected from its power source. Since then, Flash memory has evolved into the preferred storage media for a variety of consumer and industrial devices.

Flash memory is widely used in consumer devices, such as notebook computers, digital cameras, PDAs, mobile phones, GPS, electronic musical instruments, television set-top boxes, MP3 players, etc.

Flash memory is also used in many industrial applications where reliability and data retention in power-off situations are key requirements including, military systems, solid-state disk drives, networking and communication products, wireless communication devices, retail management products, medical products, etc.

Flash memory devices are used in end systems for several purposes including:

• To store and allow the processor to directly execute software code or instructions that enable customer products to boot up and operate after the power is switched on;
• To store data such as text or multimedia files; and
• To store a mixture of both code and data.

Principle of operation

Flash memory is a type of solid-state technology, which basically means that there are no moving parts. Inside the chip is a grid of columns and rows, with a two-transistor cell at each intersecting point on the grid. A thin oxide layer separates the two transistors. One of the transistors is known as the floating gate, and the other one is the control gate. The floating gate's only link to the row, or wordline, is through the control gate. As long as this link is in place, the cell has a value of '1'.

Electronic memory comes in a variety of forms to serve a variety of purposes. Flash memory is used for easy and fast information storage in such devices as digital cameras and home video game consoles. It is used more as a hard drive than as RAM.

A fundamental capability of Flash memory is to "remember" after electrical power is shut off, unlike DRAM and SRAM technologies that lose their contents when the power is removed. Flash memory is described as a non-volatile memory while DRAM and SRAM are described as volatile memories.

There are several different types of memory cards available including Compact Flash, SmartMedia, and Memory Stick to name a few. Memory cards come in a range of storage capacities starting as low as 2MB to as much as 4GB. For the most part, these memory cards use a type of memory called Flash Memory (sometimes called Flash RAM). It is a type of non-volatile memory that can be erased and reprogrammed in units of memory called blocks. Flash memory derives its name because the microchip is organised so that a section of memory cells are erased in a single action or "flash."

Flash memory evolved from Electrically Programmable Read Only Memory (EPROM) technology originally invented in the 1970s. While EPROM only allows one location at a time to be erased or written, Flash allows multiple memory locations to be erased or written in one programming operation. This allows Flash devices to be erased and re-programmed "in system", which has opened an enormous range of new applications. Note that the Flash memory erase requirement is fundamentally different from volatile memory technologies such as DRAM and SRAM where previously written contents may be overwritten without being erased first.

Jeremy Werner, solutions delivery marketing manager, wireless systems division, Spansion, said: "Spansion invented its MirrorBit technology in the late 90's and released the first product based on MirrorBit in 2001. The MirrorBit technology is revolutionary in that it allows two bits to be stored in a single transistor and later read back in a digital manner. This differs from multi-level cell Flash by physically separating the charges used to store the two bits resulting in improved reading and programming performance."

There are two major technologies of Flash memory: NOR and NAND. Each technology has strengths that make it ideal for different kinds of applications, as summarised:

NOR Flash memory: NOR, named after the specific data mapping technology (Not OR), is a high-speed Flash technology. NOR Flash memory provides high-speed random-access capabilities, being able to read and write data in specific locations in the memory without having to access the memory in sequential mode. Unlike NAND Flash, NOR Flash allows the retrieval of data as small as a single byte. NOR Flash excels in applications where data is randomly retrieved or written. NOR is most often found built into cellular phones (to store the phone’s operating system) and PDAs and is also used in computers to store the BIOS program that runs to provide the start-up functionality.

NAND Flash memory: NAND Flash was invented after NOR Flash, and is named after the specific mapping technology used for data (Not AND). NAND Flash memory reads and writes in high speed, sequential mode, handling data in small, block sizes. NAND Flash can retrieve or write data as single pages, but cannot retrieve individual bytes like NOR Flash. NAND Flash memory is commonly found in solid-state hard drives, audio and video Flash media devices, television set-top boxes, digital cameras, cell phones (for data storage) and other devices where data is generally written or read sequentially.

For example, most digital cameras use NAND-Flash based digital film, as pictures are usually taken and stored sequentially. NAND-Flash is also more efficient when pictures are read back, as it transfers whole pages of data very quickly. As a sequential storage medium, NAND Flash is ideal for data storage. NAND Flash memory is less expensive than NOR Flash memory, and can accommodate more storage capacity in the same die size. Flash memory which stores a single bit per cell is known as Single-Level Cell (SLC) Flash.

NAND memories traditionally referred to the way the transistors that store the bits are structured. Now because of the introduction of products like SuperAND and ORNAND, NAND often simply refers to the interface of a chip designed for mass data storage.

Generally, in comparison to NOR Flash, NAND Flash has faster erase and write times, but slower read times and is not randomly accessible for code execution purposes. NAND devices, because of the way their transistors are structured, are much less reliable; it is not uncommon for several percent of the memory cells of a device to fail during normal usage. Because of NOR's fast read speed, ability to be accessed randomly, and high reliability, it has dominated code storage applications.

NAND has traditionally been used for data storage. This particularly includes storage of data such as MP3 songs and pictures that are less impacted by bit errors and can be written and read sequentially.

"Going forward, we believe the market can and should focus less on the NOR vs. NAND segmentation, and more on the specific needs of applications requiring code storage, data storage or a combination of the two. Spansion's MirrorBit ORNAND is a groundbreaking alternative to traditional NAND Flash memory. The architecture offers the lower cost structure of NAND at the appropriate densities for the cell phone market, plus the reliability and fast read performance the handset market requirements. MirrorBit ORNAND was designed to specifically meet the needs of the wireless market, and complements NOR Flash memory in eXecute In Place (XIP) architectures. In code shadowing applications, MirrorBit ORNAND features superior read speeds over NAND, yielding improved user experiences vs. using NAND in those applications," added Werner.

As per a survey conducted by a leading brand in memory card solutions in India, MCC emerges as the undisputed leader amongst the branded segment in the Indian memory cards industry in terms of market share with projections of 25 percent market share by fiscal end. The Indian industry, valued at 18,00,000 cards a year in volume and Rs. 180 crores in value, has MCC's growth projections of a 20 percent market share with approximately 3.75 lakh cards in volume.

Types of memory cards

The PC Card (also known as a PCMCIA card) is one of the earliest types of memory cards that were used in digital cameras. The PC Card was originally developed to provide expandable system memory for portable computers, and was also a media used to load programs on to early portable computers. Because this type of memory card is fairly large, its use limited the ability of camera makers to produce small digital cameras, The PC Card also used a lot of power. Cameras that used these memory cards had to have large batteries and in addition, used the batteries up very quickly. The PC Card is a very common type of card, which is still used in laptop computers as a means to add functions such as modems, Ethernet and SCSI interfaces, and for transferring large files between computers. It also tends to be very durable, and is well shielded from electromagnetic fields by it's metal casing.

The PC Card comes in three different types - Type I, Type II and Type III. The Type I and II cards are solid-state type memory cards, while the Type III cards are usually tiny hard drives. All these card types measure the same length and width and use the same 68-pin connector. The only difference between card types is thickness. The thickness is 3.3, 5.0, and 10.5 millimeters for Type I, Type II, and Type III cards, respectively. Because they differ only in thickness, a thinner card can be used in a thicker slot, but a thicker card can not be used in a thinner slot. The Type III cards have been slowly disappearing from the market over the past few years or so and it is currently difficult to find these cards in retail outlets.

The CompactFlash memory card was developed soon after the PC Card, and in many ways is simply a smaller version of the PC Card. This was the first type of memory card that was developed with the intention of being used in digital cameras. The compact flash memory card has all of the same advantages as the PC Card, but because it is smaller, allows camera makers to design more compact digital cameras. The CompactFlash Card has several downsides, including a fairly high cost of approximately 65 cents per megabyte and relatively slow data transfer rates. However, the CompactFlash Card is now available with higher data transfer speeds allowing for faster data transfer to the computer through a card reader, and better performance form many digital cameras - especially the professional models. High Speed cards are currently available form brands such as Lexar and Sandisk at a cost typically around 70 cents per megabyte. The CompactFlash card is currently the most common type of memory card used in digital cameras, and is compatible with all current professional digital cameras.

The SmartMedia memory card is a more recent product that became commonly available in the mid 1990's. This memory card is smaller and much thinner than the Compact Flash card. Due to it's smaller size, camera makers were able to begin developing even smaller digital cameras that were similar in size to some of the smallest point and shoot film cameras. The SmartMedia Cards are also more efficient with regards to power drain, and the batteries in the digital cameras that use this type of card tend to last a bit longer. Another advantage of this card type is that they are a bit less expensive to produce than CompactFlash cards with a typical cost of under 60 cents per megabyte. There are some drawbacks to the SmartMedia card system. One problem is that the cards have no shielding, so they are susceptible to data loss and damage from electromagnetic fields (such as airport x-rays). Also, these cards are not as widely compatible as the other types of memory cards - Particularly CompactFlash cards. CompactFlash cards have a control circuit built right into the memory card. This circuit allows the card to "tell" the camera or computer how much capacity it contains so the camera or computer can take advantage of all the available memory. The SmartMedia Cards do not have this controller circuit, so the memory controller needs to be built into the camera. What this means to the user, is that the camera you purchase will not be able to read any cards of larger capacity than was available when the camera was released. For example, say you purchased an Olympus D-600L camera. This camera was released in early 1998 when the largest SmartMedia card available was 16MB. Because the Controller chip in the camera was designed for 16MB cards, you cannot use cards larger than 16MB capacity in this camera - larger capacity cards will not be recognised by the Olympus D-600L camera.

The Memory Stick Memory card was developed by Sony and was first used in Sony digital cameras in 1998. The Memory Stick memory card is well shielded and durable like the Compact Flash cards but is much smaller to allow it to be used in smaller products such as ultra compact digital cameras, and MPEG players. At this time, the Memory Stick is used mostly with Sony products, but there are now several digital photo products from other manufacturers that accept the Memory Stick cards. There is also a version of this card with copyright protection technology called the MagicGate Memory Stick. This card is a bit more expensive than the standard Memory Stick card, but that additional cost pays for the potential royalty fees when recording copyrighted MPEG music.

The MultiMedia Card (MMC) was developed to try and incorporate the best features of SmartMedia and Compact Flash cards in to one system. This memory card system is very similar to the Memory Stick card with regard to its capabilities but is a non-proprietary form. This card is fairly common in products such as ultra compact digital cameras, PDAs, cell phones and MPEG players. These cards cost about the same as the Compact Flash type cards at a typical cost of slightly more than 60 cents per Megabyte.

The Secure Digital (SD) Card is similar to the MMC in appearance and is cross-compatible with the MultiMedia card. It is a highly secure stamp-sized flash memory card with cryptographic security for copyrighted data based on proven security concepts from DVD audio. Jointly developed by Matsushita Electronic (best known for its Panasonic brand name products), Sandisk and Toshiba, the SD Card maintains a high transfer rate for fast copy/download.

Developed by Fujifilm and Olympus, and manufactured by Toshiba Corporation, the xD-Picture Card is a high-capacity flash memory format for digital cameras. The xD-Picture Card is currently the smallest storage format of its kind on the market - comparable in size to a penny, measuring just 0.79" x 0.98" x 0.07" and weighing less than one-tenth of an ounce - with the potential for up to 8GB storage capacity in a single card. Currently though, the xD-Picture card is available in capacities up to 128MB, with a 256MB card planned to be available soon. This unprecedented combination of small size and large capacity will allow Fujifilm and other companies to manufacture sleeker and lighter digital cameras than ever before. Both Fujifilm and Olympus have begun to offer cameras compatible with the xD-Picture Card.

If you currently own a notebook computer, you probably have a card reader already. It is the PC card slot that is usually found on the side of your notebook computer. This card slot is usually for PC Card Type II or Type III cards and is often where the modem or network interface is installed. To use this slot as a card reader, you will need to purchase an appropriate card adapter and insert your memory card with the adapter into the card slot. When the computer recognizes the memory card, it should show up as a hard drive. Then, all you need to do is simply drag and drop your image files to a folder on your computer hard drive, and you are done. The PC Card slot is usually much faster than the camera interface or a card reader for transferring the files. Card adapters are available for just about every memory card type on the market.

There is one other option that is available for SmartMedia, MultiMedia and Memory Stick type memory cards; it is the FlashPath Floppy Disk Adapter. This device allows you to transfer the files from your memory card using the floppy disk drive on your computer. The FlashPath adapter looks like a floppy disk. You insert the memory card into the adapter and then put the adapter into your floppy disk drive. You can now view the files on the floppy drive of your computer and drag-and drop the files to your hard drive. This device does not usually work on laptop computers or external floppy drives, and is very slow. However, for the computer novice, it is the easiest way to get the files in to your computer.

Flash memory over HDDs

According to Ajay Mehta, CEO, MCC Group, Flash memory has limitations on various fronts, when compared to HDD. "Flash Memory cards are available up to only 4GB space, whereas HDD are available up to 80GB. The chances of memory cards getting corrupt are higher as compared to HDD. In addition, memory cards are small in size in comparison to HDD, which makes memory cards easy to carry but are delicate and requires extra care while carrying," he added.

"Flash memory used as a hard drive actually has many advantages over a traditional hard drive. It's non-volatile or solid state, meaning there are no moving parts. This is important because with moving parts there is a greater propensity for failures when a device is dropped. Additionally, with no moving parts, the power consumption is reduced, enabling improved battery lives for portable products. With it's improved power consumption, there is no need for a fan, and with no moving parts in the system, the device is silent. Additionally, Flash is much smaller than a traditional hard drive, allowing smaller end product form factors. Finally, Flash can be read orders of magnitude more quickly than an HDD reducing the requirements of RAM in the system and improving the speed that files load," noted Werner.

"Flash memory can lead to further miniaturisation of host devices, such as cameras, MP3 players, etc. and longer battery life compared to miniature HDDs. Flash memory will also be used in future computer platforms to speed up computer boot-up times and supplement computer memory as needed," added Dave Lee, digital media product manager, APAC business division, Kingston.

The increasing storage capacities and declining costs of manufacture have led to the steady growth of Flash memory as the memory of choice for portable products, from mobile phones, PDA's and digital cameras to digital music players. Flash memory offers random access of data on the Flash card, meaning that any location on the card can be read in one action, with very low power consumption. Traditionally, Flash memory sales have been dominated by products that store one bit of information per cell. This single bit technology is used in both NOR and NAND devices using floating gate technology. Spansion's MirrorBit technology uses a different approach to storing charge than floating gate technology by storing two bits per cell.

Commenting on the advantages of MirrorBit technology, Werner said: "MirrorBit offers manufacturing cost advantages beyond the die size reduction inherent in a two bit/cell technology. Compared with floating gate technology, MirrorBit has 10 percent fewer overall mask steps, and 40 percent fewer critical mask steps. This drives faster yield learning and overall higher yields. We believe this results in a one-process node generation cost advantage vs. competitive MLC floating gate technology. Additionally, because of the way the cells are structured, the transistors can be packed more densely on a piece of silicon than when using Floating gate technology."

Need to focus on product development, system support

The Flash market is currently divided in two major categories, based on application "integrated" storage where applications are designed into the product and "removable" storage such as USB sticks, camera cards and hard disk replacement. Spansion's focus is primarily on the "integrated" market, developing products that are specifically designed to meet the needs of wireless or embedded end products. "We see a trend where companies need to focus their product developments and system solution support toward the specific end-product markets they are targeting. The requirements of the "integrated" market are for more system support, simplified design, highly reliable products and improved to market," said Werner while commenting on the latest trends in the Flash memory market.

"Cell phone manufacturers have always been the most aggressive adopters of low-voltage devices due to the importance of battery life. This pushes the voltage of Flash memory to 1.8V and adoption of 1.8V Flash memory will continue going forward. Over the last few years cell phone designers have shrunk form factors and integrated more functionality into the phone, e.g. Bluetooth, video, music. These manufacturers have increasingly demanded that Flash devices be stacked in a single, multi-chip package along with the RAM (MCP) to help reduce the form factor of handsets. We think this trend for integration will continue to develop and we have been very involved in the development of package-on-package (POP) to integrate the memory subsystem with the baseband and application processors in the phone," noted Werner.