- Show more
The evolution and history of flash memories
This document exposes the rapid evolution of Flash memories from the last decades of the twentieth century to the present. This type of storage mode, a direct heir of the EEPROM memories is characterized by writing or reading several memory cells in a single operation, as opposed to its predecessor that only allowed reading a single cell by operation by operation. This technology was largely popular for its application in digital photography where it revolutionized the way of storing information. After currently arriving at the Flash Memoirs, they are integrated into USB Memoirs and SSD technology.
The scope of this report is located in the computer science, to be more concrete in the field of hardware and information storage systems. It is an issue on which there are numerous works and analysis either developed by companies that market these components or by research entities such as universities, for this reason it is an issue that has a wide theoretical basis and state of the art reasonably developed. The objective of this report is to analyze the development and evolution of flash memories as well as the paradigms that open for their future progress. In the next lines, the precedents that laid the basis of this technology will be seen first, then the electronic foundations will be analyzed and to finish its commercialization and the future of said hardware will be analyzed.
Although Flash memories are currently a true technological family, in origin they were nothing more than the evolution of its precedent, the EEPROM "Electically Programmable Electrical Read-Only Memory MEMORY" MEMORY MEMORY ". The EEPROM are a non -volatile type of memory capable of erase and reprogram their memory cells applying electric voltages instead of using laser like other types of memories. They are characterized by having no limitation regarding the number of occasions that can be read, but it does have it in terms of erase and reprogramming, being its useful life currently around 100.000-1.000.000 clock cycles. Attending to the technical details, their memory cells are formed by silicon mos transistors with floating door which is called Samos Structure. At the time when the first flash memories were created the EEPROM only allowed to operate on a by cycle but are currently already able to operate on several bytes in a single clock cycle. This same technical feature was what motivated the development of flash, since the fact that the EEPROM only acted on a byte made them considerably slow. For this reason the main advantage of flash memories over the EEPROM consists of their speed, which is far superior to being able to act on bytes blocks.
DEVELOPMENT AND COMMERCIAL LANGE
The creation of Flash technology is rigged with the name of Fujio Masuoka. This man was a worker of the Japanese Toshiba company in which he had invented Samos memory in 1971, which allowed the creation of the EEPROM. Later he investigated and patented in 1980 the floating door technology, fundamental in the flash memories. As for the anecdotal, it is worth mentioning the reason for the name "Flash" that comes from an appreciation of a Masuoka colleague called Shoji Ariizumi: "His colleague Shoji Ariizumi suggested the word" flash "because the erased process reminded him of the flash of acamera". In 1983 the invention of the flash-nor was patented, created by the same Masuoka team but it would not be until 1987 when the flash saw their launch against the general public. This happened in an electronic devices congress that took place in San Francisco in which Toshiba presented the flash-nand for its subsequent commercial launch.
Although Toshiba was the company that financed and carried out the investigations and experimentation that would lead to the creation of flash, really the company that showed off the discovery and took advantage of the potential of this technology was the American company Intel, which, whichIt would hardly take a year to respond to the launch of flash-nand with their commercialization of flash-nor. But it would not be until 1994 when the true flash revolution would occur. This occurred following the creation of the first memory cards by the Sandisk company. The great utility and application in the real life of flash was appreciated by society with its use in digital photography. People, accustomed to analog photography in which only 24-36 photographs could be stored per reel (usually) was impressed to see as a simple card of a very small size was able to save hundreds even thousands of snapshots. Compactflash compact flash were not only used in photography but in many other areas such as audio reproduction or video game industry.
Currently, the use that is usually given to Flash memory is on cards for storage of photographs and videos, the popular USB memories and from 2009 approximately in the solid state units that today are promoting enormous progress in storage systemsand imposing itself to its direct competitors, the HDD discs.
Fundamentals of flash memories
As indicated before, the EEPROM memories were based on the Samos structure, instead flash memories use another type of transistor called Famos Transistor, “Floating-Gate Avalanche-Injection MOS”, which like the Samos structure uses the systemfloating door. This system is characterized by adding an additional door to the existing door between drain and source. In the case of the famous transistor the floating door is usually made of polysilicio and is surrounded by an insulating material. A negative voltage is applied to the transistor that causes electrons to appear on the surface close to the two existing doors. Through a phenomenon known as tunnel effect it is possible to inject these electrons into the floating door. When the voltage that was previously spoken (programming voltage) is stopped apply. The deletion will be achieved by releasing the enclosed electrons and this happens when the Fowler-Nordheim tunnel effect is induced. The characteristics of the polysilicio allow to store the information up to about 20 years, time that the floating door can be kept.
Not all flash devices are exactly the same because there are two types, those based on nor doors and the nand doors. Although its fundamental characteristics (non-volatile memory that allows you to read and write about blocks of bytes) are the same there are implementation details that make them behave differently. The transistor density of the NAND is much greater than that of the NORs in addition to the cost of the NAND is remarkably smaller. As for reading speeds, they are superior in the NORs and in writing appears the great fort of the Nor because with a Nor flash we can modify an individual byte instead if we want to make any change with a nand we would have to erase the entire block,modify the desired byte and leave the rest of the memory with its previous state. In addition, NORs are said to be much more reliable because the possibility of data corruption is very small. This means that according to the needs that will be used, one type of flash or other will be used. For very delicate uses and in which a small error (induced by corruption) can cause a very adverse result the decision will be clear and a nor will be chosen. On the other hand, for more general uses such as the needs of the general population, see, for example, USB memories will be chosen, as they are much cheaper and have greater densities.
As previously seen the flash memories are an evolution of the EEPROM that due to their characteristics they have managed to settle in today’s society and they are not expected to be abandoned in the near future. The counter is expected to continue evolving and improving, Moore’s law in non-volatile memory continues to be fulfilled although it is predicted that this law will soon stop being valid when reaching the physical limits presented by transistors. A solution to this is already being investigated and it is about using graphene instead of silicon for the manufacture of transistors. This could be a huge revolution in the event that it could be carried out in a massive way. With which in the future we could have graphene -based flash memories. More in the short term the potential of flash goes through being a component of solid state devices, which are already being incorporated into the vast majority of desktop and laptops computers. The SSD stand out not for their storage capacity but for their transfer speeds and are winning the battle to HDD mechanical discs.
This report could be expanded in different ways. The possibility of studying in greater detail the functioning of famous transistors as well as analyzing what is the way in which flash memories within SSD devices are implanted.