March 11, 2016 at 15:07The history of the development of flash memory
Hi Giktayms!All of us, to one degree or another, use digital technology, but not all of us think about how this technology works in this way, and not otherwise. In this article, I will talk about the integral part of almost any device - flash memory. This component is used wherever possible: we encounter flash memory hundreds of times a day, without realizing it ourselves. Flash memory is used both in portable gadgets (laptops, smartphones, players, watches), and in stationary electronics (TVs, PCs, monitors and even washing machines). But if you ask an ordinary person about flash memory, first of all he will name the most obvious things: SD- and microSD-memory cards, flash drives and the like. In fact, it is a competitor to traditional hard drives (HDDs),
Habr old-timers will surely remember the informative material about the origin of SSD and flash memory. But if the "solid-state" were invented at the turn of the century, then ordinary flash memory was born much earlier than many of you might think. Back in the middle of the 20th century, ballistic scientist Wen Tsing Chow in one of the divisions of American Bosch Armaset out to improve the memory blocks of the coordinates of the on-board computer of the Atlas E / F rocket system. But even then, the technology that came to light remained secret for several years, since this organization worked for the US government, and the existence of the Atlas system was not widely publicized. But a little time passed, and most of the company's developments were declassified, and with it the technology that Wang Qing Chow developed - it was called PROM (programmable read only memory).
The principle of PROM was quite simple: the memory was a coordinate grid, at the nodes (intersections) of which the conductors were closed with a special jumper. At that moment when it was necessary to determine the state (value) of the cell with certain coordinates, it was enough to find out if there is current at the intersection of the necessary conductors.
So the beginning has been made. But the disadvantage of the new invention was too small a capacity. However, EPROM also had advantages - high speed and resistance to mechanical damage.
The next major advancement in the development of flash memory was made by Dov Frohman from Intel. Studying the damage to failed circuits, he invented a new standard for EPROM memory . In the novelty, a special electron movement along the chip was used to save information, and powerful ultraviolet radiation was used to erase the data. EPROM memory can be found in BIOS chips on computers at the beginning of the IT era, for example, on the ZX Spectrum , popular in those years in the West , better known in the CIS under the names of Search or Magic. Intel
Guysdecided not to stop there, and in 1978, on behalf of engineer George Perlegos, the company introduced the Intel 2816 chip, similar in characteristics to EPROM, but thanks to the thin insulation layer, the “2816th” could erase information without the help of ultraviolet rays. This architecture was named EEPROM (Electrically Erasable Programmable Read-only Memory). But she had one very significant problem: because of the difficulties in implementing the correct supply of current to such a thin dielectric layer, EEPROM was deprived of the possibility of rewriting (in other words, it was “one-time”).
As a result, the developers decided to create two types of microchips based on EEPROM: the first one had a large capacity, but could not be rewritten, and the second was rewritable, but contained less information.
The solution to this problem was found by engineer Fujio Masuoka from Toshiba. To his colleague, the process of erasing data seemed like a flash - that is how the name flash-memory appeared. The innovation was presented to the public in 1984, in 1988 Intel introduced the first commercial samples of NOR-flash memory, and in 1989 Toshiba announced the usual NAND memory for us.
Information-saving circuits are called Single-Level-Cell (SLC) . Along with SLC, alternative microcircuits appeared, accommodating 2 bits of information - Multi-Level-Cell . MLC-chips turned out to be cheaper in production, but they worked slowly and were short-lived. In recent years, eMLC-memory (Enterprise Class MLC)capable of opposing MLC chips with a higher read and write speed, as well as an extended life. In terms of price / quality ratio, modern eMLC options are only slightly inferior to SLC, but at the same time they cost half as much. If you remember, last year we talked in detail about the device chips.
But back to the flash memory itself. Its operation is impossible without service chips responsible for managing the storage of information on NAND chips. These service chips were called FSP (Flash Storage Processor). Modern FSPs have great computing power: they consist of two cores, have the ability to provide advanced data protection and manage the processes of finding and destroying digital “garbage”. Such multifunctional microcircuits have rather long development periods, and the chips currently being developed will be used in production only after 2-3 years (as, for example, in many modern SSDs). By the way, SSDs are the most interesting direction for manufacturers of devices based on flash memory. The reasons are quite obvious: This is a combination of high capacity combined with an unusually high speed. Solid-state drives increase their volume literally every year, and if just a couple of years ago the average SSD's was able to hold 128-256 GB, now this figure is close to a terabyte.
If we talk about the degree of development of various options for the use of flash memory, then SSD disks already have the intended development path and quite quickly reach perfect characteristics. Things are much more interesting with memory cards: although they are not “lagging behind,” the potential for increasing speed and capacity is more impressive. To understand why this happens, let's talk about the appearance of the most common formats of removable flash drives.
It was 1999, when the companies SanDisk, Toshiba and Matsushita (now known as Panasonic) cooperated and decided to create a new unified standard for memory cards, called SD, or Secure Digital. It was on the word Secure that the main emphasis was made in the name of the new standard - cards received DRM support, or, more simply, with their advent, digital copyright protection became possible. Already in the first year of their existence, the union of the three above-mentioned companies, they created the SD Association, which one by one attracted new members, including such giants as Intel, Kingston, Apple, AMD, Canon, Nikon, Samsung, Hewlett- Packard and many others. The maximum capacity of the first models of SD cards was only 2 GB, but very soon 4 GB options appeared,
At first, this size was quite enough, but the digital industry did not stand still, the volumes of content grew by leaps and bounds, and at some point, users began to feel a clear lack of free space. Therefore, in 2006, the second generation of SD drives was introduced, called SDHC (Secure Digital High Capacity, or SD cards with high capacity); their maximum volume has grown to 32GB. This format had only one serious drawback - the lack of backward compatibility, that is, they were "not friends" with old card readers. But then again, as time went on, user requests grew, and increased the need for drives of greater capacity. Thus, SD cards version 3.01, or SDXC - Secure Digital eXtended Capacity. New to this day,
In the 2000s, following the rapid development of mobile technology, the need for more memory came. By analogy with the computer world, microSD cards began to be used in mobile gadgets and other portable devices (from a GPS navigator to headphones). This format is still popular today, and its characteristics are 4 times smaller in physical size than SD-drives, no worse than similar indicators of adult memory cards. Also, drawing parallels with SD cards, there are microSDHC and microSDXC formats that offer similar volumes of information.
At the dawn of the mobile era, quite often it was possible to meet the following formats, which are not in great demand today: Compact Flash (1994, had the highest data transfer rate at that time - up to 90 MB / s), Memory Stick (1998, the maximum amount was 128 MB), Memory Stick Pro (announced in 2003, the maximum capacity is up to 4 GB, Memory Stick Duo (2003, the first cards were equal to half the usual MS), Memory Stick Pro Duo (2006, cards reached 32 GB), Memory Stick HG-Duo (2008, the format was the last iteration of MS Duo-cards).
The history of memory cards cannot be represented in the form of direct time - it is a complex, complicated scheme. Some formats took root and began to evolve, while others turned out to be useless to anyone. In the next part of the article we will talk about more exotic memory formats and the reasons for their unpopularity.
Thanks for watching, and stay with Kingston on the Guktime!
For more information on Kingston and HyperX products, visit the company's official website . A page with a visual aid will help you choose your HyperX kit .
How it all began
Habr old-timers will surely remember the informative material about the origin of SSD and flash memory. But if the "solid-state" were invented at the turn of the century, then ordinary flash memory was born much earlier than many of you might think. Back in the middle of the 20th century, ballistic scientist Wen Tsing Chow in one of the divisions of American Bosch Armaset out to improve the memory blocks of the coordinates of the on-board computer of the Atlas E / F rocket system. But even then, the technology that came to light remained secret for several years, since this organization worked for the US government, and the existence of the Atlas system was not widely publicized. But a little time passed, and most of the company's developments were declassified, and with it the technology that Wang Qing Chow developed - it was called PROM (programmable read only memory).
The principle of PROM was quite simple: the memory was a coordinate grid, at the nodes (intersections) of which the conductors were closed with a special jumper. At that moment when it was necessary to determine the state (value) of the cell with certain coordinates, it was enough to find out if there is current at the intersection of the necessary conductors.
First successes
So the beginning has been made. But the disadvantage of the new invention was too small a capacity. However, EPROM also had advantages - high speed and resistance to mechanical damage.
The next major advancement in the development of flash memory was made by Dov Frohman from Intel. Studying the damage to failed circuits, he invented a new standard for EPROM memory . In the novelty, a special electron movement along the chip was used to save information, and powerful ultraviolet radiation was used to erase the data. EPROM memory can be found in BIOS chips on computers at the beginning of the IT era, for example, on the ZX Spectrum , popular in those years in the West , better known in the CIS under the names of Search or Magic. Intel
Guysdecided not to stop there, and in 1978, on behalf of engineer George Perlegos, the company introduced the Intel 2816 chip, similar in characteristics to EPROM, but thanks to the thin insulation layer, the “2816th” could erase information without the help of ultraviolet rays. This architecture was named EEPROM (Electrically Erasable Programmable Read-only Memory). But she had one very significant problem: because of the difficulties in implementing the correct supply of current to such a thin dielectric layer, EEPROM was deprived of the possibility of rewriting (in other words, it was “one-time”).
As a result, the developers decided to create two types of microchips based on EEPROM: the first one had a large capacity, but could not be rewritten, and the second was rewritable, but contained less information.
The solution to this problem was found by engineer Fujio Masuoka from Toshiba. To his colleague, the process of erasing data seemed like a flash - that is how the name flash-memory appeared. The innovation was presented to the public in 1984, in 1988 Intel introduced the first commercial samples of NOR-flash memory, and in 1989 Toshiba announced the usual NAND memory for us.
Information-saving circuits are called Single-Level-Cell (SLC) . Along with SLC, alternative microcircuits appeared, accommodating 2 bits of information - Multi-Level-Cell . MLC-chips turned out to be cheaper in production, but they worked slowly and were short-lived. In recent years, eMLC-memory (Enterprise Class MLC)capable of opposing MLC chips with a higher read and write speed, as well as an extended life. In terms of price / quality ratio, modern eMLC options are only slightly inferior to SLC, but at the same time they cost half as much. If you remember, last year we talked in detail about the device chips.
But back to the flash memory itself. Its operation is impossible without service chips responsible for managing the storage of information on NAND chips. These service chips were called FSP (Flash Storage Processor). Modern FSPs have great computing power: they consist of two cores, have the ability to provide advanced data protection and manage the processes of finding and destroying digital “garbage”. Such multifunctional microcircuits have rather long development periods, and the chips currently being developed will be used in production only after 2-3 years (as, for example, in many modern SSDs). By the way, SSDs are the most interesting direction for manufacturers of devices based on flash memory. The reasons are quite obvious: This is a combination of high capacity combined with an unusually high speed. Solid-state drives increase their volume literally every year, and if just a couple of years ago the average SSD's was able to hold 128-256 GB, now this figure is close to a terabyte.
If we talk about the degree of development of various options for the use of flash memory, then SSD disks already have the intended development path and quite quickly reach perfect characteristics. Things are much more interesting with memory cards: although they are not “lagging behind,” the potential for increasing speed and capacity is more impressive. To understand why this happens, let's talk about the appearance of the most common formats of removable flash drives.
SD cards
It was 1999, when the companies SanDisk, Toshiba and Matsushita (now known as Panasonic) cooperated and decided to create a new unified standard for memory cards, called SD, or Secure Digital. It was on the word Secure that the main emphasis was made in the name of the new standard - cards received DRM support, or, more simply, with their advent, digital copyright protection became possible. Already in the first year of their existence, the union of the three above-mentioned companies, they created the SD Association, which one by one attracted new members, including such giants as Intel, Kingston, Apple, AMD, Canon, Nikon, Samsung, Hewlett- Packard and many others. The maximum capacity of the first models of SD cards was only 2 GB, but very soon 4 GB options appeared,
At first, this size was quite enough, but the digital industry did not stand still, the volumes of content grew by leaps and bounds, and at some point, users began to feel a clear lack of free space. Therefore, in 2006, the second generation of SD drives was introduced, called SDHC (Secure Digital High Capacity, or SD cards with high capacity); their maximum volume has grown to 32GB. This format had only one serious drawback - the lack of backward compatibility, that is, they were "not friends" with old card readers. But then again, as time went on, user requests grew, and increased the need for drives of greater capacity. Thus, SD cards version 3.01, or SDXC - Secure Digital eXtended Capacity. New to this day,
microSD: towards the mobile element
In the 2000s, following the rapid development of mobile technology, the need for more memory came. By analogy with the computer world, microSD cards began to be used in mobile gadgets and other portable devices (from a GPS navigator to headphones). This format is still popular today, and its characteristics are 4 times smaller in physical size than SD-drives, no worse than similar indicators of adult memory cards. Also, drawing parallels with SD cards, there are microSDHC and microSDXC formats that offer similar volumes of information.
At the dawn of the mobile era, quite often it was possible to meet the following formats, which are not in great demand today: Compact Flash (1994, had the highest data transfer rate at that time - up to 90 MB / s), Memory Stick (1998, the maximum amount was 128 MB), Memory Stick Pro (announced in 2003, the maximum capacity is up to 4 GB, Memory Stick Duo (2003, the first cards were equal to half the usual MS), Memory Stick Pro Duo (2006, cards reached 32 GB), Memory Stick HG-Duo (2008, the format was the last iteration of MS Duo-cards).
To be continued
The history of memory cards cannot be represented in the form of direct time - it is a complex, complicated scheme. Some formats took root and began to evolve, while others turned out to be useless to anyone. In the next part of the article we will talk about more exotic memory formats and the reasons for their unpopularity.
Thanks for watching, and stay with Kingston on the Guktime!
For more information on Kingston and HyperX products, visit the company's official website . A page with a visual aid will help you choose your HyperX kit .