Random Access Memory

All computers contain what is called RAM, which stands for Random Access Memory. RAM is a form of computer data storage, which takes the form of integrated circuits. An integrated circuit is a smaller version of an electronic circuit made of semiconductor material. This allows the data to be accessed at random, hence random access memory. RAM is most often associated with volatile types of memory such as DRAM, which stands for Dynamic Random Access Memory. In fact, most types of RAM used in modern computers are volatile. With this type of memory, any information on the computer is lost when that computer is powered down. With non-volatile memory, information is retained even after the computer is powered down.

The most common type of RAM found in personal computers and workstations is DRAM. Unlike Static RAM (SRAM), DRAM requires its storage cells to be refreshed every few milliseconds. It operates on the principle that currents move in one direction or another, whereas SRAM operates on the principle that storage cells will hold the charge in place. The way DRAM works is by sending a charge through a column, which will activate the transistor at each bit contained in the column. DRAM contains memory cells that have a transistor and capacitor paired together, which is why they must be refreshed constantly.

In 1966, Dr. Robert Dennard invented DRAM while working at the IBM Thomas J. Watson Research Center and was awarded a U.S. patent for it in 1968. In 1969, Honeywell, a major conglomerate company, requested that Intel make a DRAM, which would use a 3-transistor cell that Honeywell had developed. In 1970, it became the Intel 1102. When it was discovered that the 1102 had many problems, Intel came up with an improved design, which became the Intel 1103, released in 1970. The Intel 1102 never made it to the computer market. The Intel 1103 came to be the first commercially available DRAM memory. It was the worlds first 1K dynamic RAM and marked the turning point in the integrated circuit’s history. With the Intel 1103, a significant amount of information could now be stored on one chip. Its capacity to hold a large amount of data led to the Intel 1103 replacing core memories and it rose to become the industry standard. The Intel 1103 became the largest selling semiconductor in the world in 1972.

Core memory, which is also known as magnetic core memory, is a RAM system developed at MIT in 1949 by Jay Forrester, a pioneer in the early development of digital computers and inventor of random-access, coincident-current magnetic storage. It went on to become the dominant form of memory in the 1950’s and was used until the late 1970’s. Core memory is a non-volatile memory, which includes special logic in the memory controller. This is to ensure that the location of the memory’s content cannot be altered without the voltages being at the proper levels. The purpose of this is to prevent random changes to the content of the memory when power is lost and restored.

Static Random Access Memory (SRAM) is a form of semiconductor memory. In most cases, SRAM is used for cache memory, because it can be accessed much quicker than DRAM. It is also used in video cards to provide the card with RAM digital to analog conversion. An SRAM cell can be in one of three different states; standby, reading and writing. Standby occurs when the circuit is not in motion. Reading is when data has been requested from the chip and writing is updating the contents of the chip. While SRAM is more expensive than DRAM, it is faster and uses slightly less power than DRAM does. SRAM is used when low bandwidth and/or low power is a factor. It provides an easier user interface and allows more random access than many types of DRAM. SRAM boasts a more complicated structure than DRAM and is less dense. Therefore, it is not used as the main form of memory in personal computers.

Another type of memory chip is Extended Data Output (EDO), which is a common type of asynchronous DRAM, and is sometimes referred to as hyper page mode DRAM. In addition, EDO improves read time on microprocessors such as the Intel Pentium. EDO RAM is a lot like fast page mode DRAM, which is memory that is slightly faster than DRAM. Despite the name, fast page mode (FPM) DRAM is in fact, the technology with the slowest memory in modern PCs. When it comes to manufacturing costs, EDO and FPM cost the same amount, yet EDO is so prominent in the computer market that it has become cheaper than FPM, despite being the superior memory. EDO memory is found in most of the later models of Pentium-class PCs. Support from the system chipset is required for EDO memory. Since the invention of EDO in 1994, many of the newer Pentium computers do support EDO, while older systems do not.

Another version of RAM is Burst Extended Data Out (BEDO) RAM, which is an improvement on the conventional asynchronous RAM. BEDO memory is simply EDO memory combined with pipelining technology that has special latches, which serve to provide faster access time than basic EDO. One difference between BEDO memory and EDO memory is that BEDO memory can accommodate much higher bus speeds than EDO can. BEDO memory improves the performance of DRAM, for just about the same cost to produce it. It is said that BEDO is a bigger improvement over EDO than EDO is over FPM. However, BEDO never became popular with consumers. In terms of performance, while BEDO competes with SDRAM, it cannot beat it in the market.

Most computers manufactured in the last few years that are designed to use the modern asynchronous RAM actually support FPM. It is safe to use on the majority of new computers because it does not require any unique compatibility or support. The disadvantage, however, is that FPM offers a lower performance than the majority of other memory technologies.

Examples of other commonly used types of RAM are most types of ROMs (Read Only Memory), and NOR-FlASH, which is a flash memory. Flash memory is computer memory that can be erased and reprogrammed. It is generally found in memory cards as well as USB flash drives. The purpose of flash memory is to store and transfer data between computers.

Both the NOR and NAND Flash memories were invented by Dr. Fujio Masouka when he was working for Toshiba, approximately 29 years ago. It was named flash memory because the process of erasing the memory contents is much like the flash of a camera, according to a colleague of Dr. Masouka’s at Toshiba. Dr. Masouka first presented flash memory at the IEE 1984 International Electron Devices Meeting, which was held in San Francisco, CA. The potential of the flash memory was recognized by Intel and they introduced the initial NOR type flash chip in 1988.

At the 1987 International Electron Devices Meeting, Toshiba unveiled the NAND flash, which allowed for faster erase and write times while still providing full address and data buses. In turn, this allows any location to be randomly accessed. This led to flash memory being able to replace older ROM chips that were used to store program codes that did not require frequent updates. NAND has as much as ten times the endurance of the NOR flash.

NOR flash and NAND flash are different in two ways. One is that they provide different interfaces for reading and writing memory. When it comes to reading, NOR allows random-access, while NAND only allows page access. The other is that the connections of each memory’s individual cells are different. NOR flash was developed as a more economical ROM than EEPROM as well as a more convenient rewritable ROM.

Flash memory is a type of EEPROM (Electrically Erasable Programmable Read Only Memory,) that is erased in chunks rather than as a whole. Since Flash memory is cheaper than EEPROM, it has become the main technology in cases where a significant amount of non-volatile storage, as well as solid state, is needed. For example, PDAs and digital cameras use flash memory. It is popular in the video game console market, because it is often used in place of battery-powered SRAM and EEPROMs for the purpose of storing game data.

EEPROM is a non-volatile memory type used in computers and other electronic devices for the purpose of storing small amounts of data that need to be retained when power is removed, such as during device configuration. Manufacturers such as Mitsubishi, Hitachi, Maxwell Technologies, and Samsung Electronics use EEPROM in their products.

George Perlegos designed the EEPROM chip in 1978 while working for Intel. Perlegos then left Intel and formed a company called Seeq Technology where he went on to develop the first EEPROM that was fully functional. He modified the memory chip by integrating the capacitor circuit and oscillator directly into the memory chip itself. One disadvantage of the EEPROM is that, while it can be reprogrammed, there is a limit to the number of times it can be altered. Because of this, EEPROM chips are most widely used to store configuration data, which doesn’t require frequent programming of the computer’s BIOS code. Today’s EEPROMs are able to be rewritten a maximum of a million times.

Erasable Programmable Read-Only Memory (EPROM) is a type of memory chip that has the ability to retain data when the power is cut, i.e. a non-volatile memory. Once it has been programmed, the only way to erase the data is to expose it to a strong ultraviolet light. An EPROM chip that has been programmed can retain data between 10 and 20 years and can be read as many times as needed. Dov Frohman, an engineer from Israel, invented EPROM in 1971. He developed the concept in 1970, when he was in the process of troubleshooting problems in one of the early Intel products. At this time, there were only two types of memories; RAM and ROM. EPROM has features of both these types of memories. While it is non-volatile, like RAM, it is also easy to reprogram. The EPROM was a significant addition to the personal computer industry and had the distinction of being the most profitable product used by Intel far into the 1980s.

While there are many different types of RAM, the three main types are SDRAM (Synchronous DRAM), DDR (Double Data Rate SDRAM,) and Rambus Dram (RDRAM.) SDRAM stands for synchronous dynamic random access memory. This is a term used to describe dynamic random access memory with a synchronous interface. The SDRAM waits for a signal from the computer’s internal clock before responding to any input. By doing so, it synchronizes itself with the computer’s system bus. This process means that the chips within the computer operate under a more complicated pattern of operation than DRAM, which is not synchronized with the computer’s clock.

Double Data Rate SDRAM, or DDR SDRAM, is a faster version of SDRAM. It doubles the rate of transfer over SDRAM by using a clock cycle. It uses a similar parallel bus to SDRAM, but is physically incompatible with DDR, which makes it easier to implement than RDRAM. Since the DDR2 SDRAM surpasses DDR SDRAM, it is at times, referred to as DDR1 SDRAM. DDR SDRAM began being used as memory for most PCs beginning in 2000. The original intent was to provide minor enhancements to the DDR SDRAM that boasted higher clock rates and slightly deeper pipelining. The development and introduction of DDR3 SDRAM in 2007 was anticipated to quickly replace DDR and DDR2.

DDR2 SDRAM is the offspring of SDRAM and was first introduced in 2003. However, due to latency problems, it never outperformed the original DDR. The DDR2 was competing with the standard DDR by the end of 2004.

A type of memory used in certain portable electronic devices is Mobile DDR SDRAM (MDDR.) It is used in cell phones, digital audio players, and other handheld devices. MDDR operates at a higher voltage than DDR SDRAM, allowing a reduction in power consumption.

Rambus Incorporated, which was founded in 1990, provides high-speed interface technology. The company is well known for introducing the DDR-SDRAM memory. In 1996, an agreement was reached with the Intel Corporation, which obligated them to use RDRAM, until 2002, as the main memory technology for all Intel platforms.

Rambus Dram (RDRAM) is a different technology than SDRAM. It uses a special high-speed data bus, which is called The Rambus channel. Because the RDRAM is such high-speed memory, it generates much more heat than most other types of chips. To counteract this, Rambus chips have a long thin wafer shaped heat spreader designed to keep the chip from over heating. RDRAM is the only memory that Intel uses for their Pentium 4. In 1999, the first PC motherboards with RDRAM were put on the market. It has since been used in Nintendo 64 and Playstation 2.

There are disadvantages to using RDRAM, including cost, heat output, and delays in accessing information. These disadvantages caused RDRAM to become obsolete, at which time Rambus went on to support DDR and DDR2 in relation to video card technology. Rambus also developed XDR RAM technology. Extreme data rate DRAM is simply a RAM interface boasting high performance. It is in direct competition with DDR2 SDRAM technology as well as GDDR4 technology. The XDR eliminates the problems that interfered with the earlier forms of RDRAM, such as unusually high latency. This technology is most famously used in Sony’s Playstation 3.

Graphics Double Data Rate (GDDR) refers to the video memory used in modern video cards. There are 5 versions of GDDR, which are simply referred to as GDDR2, GDDR3, GDDR4, and most recently, GDDR5. The full name of this memory is Graphics Double Data Rate Synchronous Dynamic Random Access Memory (GDDR SDRAM.) “Double Data Rate” refers to the capacity of the memory for double-pumping data. The transference of information occurs on the rising and falling edges of the memory’s clock signal. The word “synchronous” simply means the memory can operate in sync with the computer system’s bus. Through this process, the memory is able to accept new instructions while previous instructions are still processing. This is known as instruction pipelining.

GDDR2 memory never became popular due to the fact that it produces a significant amount of heat because it operates at a much higher clock speed than the original GDDR. However, GDDR3 went on to become the most common technology in graphic memory today. It was an improvement over previous versions, simply because it can accommodate higher clock speeds, using less power, than its predecessor. Another advantage of the GDDR3 is that it uses 2 separate read and write data strobes, meaning the GDDR3 boasts a much faster read-to-write ratio than GDDR and GDDR 2 did. The GDDR 3 chips also feature hardware reset, meaning the memory can be wiped clean and can begin to receive new data if necessary. The differences between GDDR 3 and GDDR 4 are nominal. The GDDR 4 simply operates with fewer volts than its predecessor. It also has an increased memory performance over previous versions. However, only AMD has chosen to use the GDDR4 technology, and because of this, the only manufacturers that use it are Samsung and Hynix. This means the price of GDDR4 memory is relatively high in comparison to other memories.

The next major development in graphics is the GDDR5. It requires fewer volts than GDDR4, which allows it to run the memory with less heat output than other memory chips. This feature of the GDDR5 has the potential to help in overclocking, as well as the ability to reduce the cost of manufacturing it. The GDDR5 will also extend the battery life when used in a notebook PC. In addition to being manufactured by Samsung and Hynix, GDDR5 is also manufactured by Qimonda.GDDR5 is about four times faster than GDDR3 and uses 20% less power. In February of 2009, Samsung began mass-producing GDDR5. It has become the fastest double-data-rate memory chip, which processes data twice as fast as the current technology, providing a big boost to the gaming world. By the end of 2010, Samsung expects GDDR5 to be found in more than half of all high end PCs.

In 1995, research began on the MRAM (magnetoresistive random access memory), which is a low power nonvolatile micro-memory cell technology. MRAM began in the 1940s at Harvard University when physicists worked with colleagues at MIT to develop the memory. Data is stored through the use of magnetic charges rather than the electrical ones used by DRAM. Magnetic polarity is used to store the data within the MRAM. The U.S. Defense Advanced Research Projects Agency funded the research to develop technology that would serve as a general-purpose memory, enabling high density, high speed, and low power consumption. MRAM technology is reportedly faster and cheaper to make than non-volatile flash memory and is said to be a better choice for desktop systems than flash memory. The advantages of MRAM over flash technology likely means that it could last indefinitely. Proponents of the MRAM believe that because of the advantages it provides that it will one day be the dominant type of memory, or a true universal memory. Much like Flash memory, MRAM is able to retain data even when the power supply is cut off. This prevents the loss of data in the event the computer is unexpectedly shut off. Flash memory requires charge pumps, which consist of electronic circuits using capacitors as elements to store energy, which create higher or lower voltage power sources. With MRAM, these charge pumps are eliminated. Because MRAMs boast extremely fast read and write times, Motorola unveiled a 1 MB MRAM chip in June of 2002.

MRAM can be found in many of today’s technological devices, such as cell phones, digital cameras, notebooks, smart cards, personal computers, media players, book readers and cb amplifier. It is also used by aerospace and military systems. Today’s developments in the MRAM chip, which boasts high data stability, allow for the commercialization of what is a low-cost, yet advanced technology chip, which will be used in disk drives and network routers among other things.

Spin Torque Transfer (STT) RAM, which is also referred to as STT-MRAM or SpinRam, is an electric current that when applied to a magnet, changes the direction of the magnetic field. The current runs up and down, and from left to right, which causes change in the resistance of the current. STT-RAM hit the market in 2008 and has surpassed MRAM as the faster chip. It replaces embedded technologies like eSRAM, eFlash and DRAM. STT-RAM provides higher speed and lower power in automotive applications than eFlash. In addition, it is denser than eSRAM. In personal computers, STT-RAM replaces SRAM for high-speed cache, as well as replacing Flash for non-volatile cache, and PSRAM as well as DRAM to execute programs at a high speed. Grandis, which is the self-proclaimed “pioneer in STT-RAM,” developed the chip in conjunction with Hynix Semiconductor Inc. In addition, the two companies collaborated to integrate this technology into the future memory products of Grandis.

An April 1, 2008, press release issued by Grandis describes STT-RAM as “a next generation non-volatile (NVM) solution that overcomes the limitations of conventional magnetic RAM (MRAM) technologies.” STT-RAM also has a high quality of scalability, leading to greater density and lower costs. In addition, many existing mainstream memories use more power than STT-RAM does. It also provides its user with fast read/write capabilities, as well as unlimited endurance. Teams from both Grandis and Hynix Semiconductor Inc. worked together to implement this technology.

One advantage of STT-RAM is that it provides a combination of the benefits of SRAM, DRAM, and Flash memory. It also offers greater scalability than most other technology chips.

In October of 2008, Grandis received $6.0 million from the Defense Advanced Research Projects Agency (DARPA) in order to fund the initial phase of research in the development of the STT-RAM chip. Because of its technological features, STT-RAM is very well suited for defense applications; hence the reason the research was funded by the DARPA.

At the forefront in the development of STT-RAM is Grandis, which led the development of the materials and structures needed to enhance the efficiency of the spin-transfer technology that is STT-RAM. In 2002, Grandis was founded in order to develop non-volatile technology through research in spintronics as well as to pursue new magnetic materials, which would lower the STT switching currents. STT is a fairly new phenomenon in physics that was first predicted in 1996 and demonstrated in 2000. Through much research and practice, Grandis enabled the development of a package that allowed the corporation of stand-alone/ embedded STT-RAM non-volatile memory into all of their memory products in the future.

In 2007, NYU and Allied Minds, which is an investment corporation that specializes in university business ventures, teamed up to start a company called Spin Transfer Technologies, LLC, in order to commercialize a more efficient form of computer memory called STT-MRAM. Researchers at NYU developed a new form of MRAM to provide non-volatile storage of critical data that is frequently updated. The STT- MRAM technology works by rapidly changing the magnetic orientation of nanometer scale magnets through the use of spin transfer technology. Allied Minds COO Marc Eichenberger expects STT- MRAM to be dominant in the next generation of technology.

In 1989, IBM discovered an important component in MRAMs development, which was the giant magnetoresistance effect (GMR.) In 1996, DARPA funded programs to begin independent research into MRAM. The companies that performed the research were Motorola, IBM, and Honeywell. In 2000, IBM combined with Infineon and formed a development program for MRAM, while in 2002; Motorola unveiled a 1MB MRAM prototype chip. By the end of 2004, IBM and Infineon had demonstrated a 16 MB prototype chip of the MRAM. In early 2006, Freescale Semiconductor was the first company to ever come out with an MRAM device. In February of the same year, Toshiba and NEC demonstrated an MRAM chip, which boasted the fastest speed ever recorded. The first commercial shipments were made in July of 2006.

In conclusion, RAM is a form of computer data storage found in all electronic devices. It is the heart of the computer system, and a key component in the storage of data. There have been many advances in computer memory over the last several decades, a trend that can only continue well into the future. The most important thing to know about RAM is that, the bigger RAM is, the faster and more efficiently the computer will run. RAM sizes vary from 128 MB all the way up to 1024 GB. The RAM on a computer can be tested by using a specialized memory test application. The subject of RAM is one that has been widely discussed among computer users and there is a surplus of information to be found about RAM on the World Wide Web. Educating yourself on everything from RAM types to the history of RAM can be achieved with a simple Internet search. RAM is a constantly evolving technology, which can only serve to boost the computer industry and allow users to perform various tasks at faster and faster speeds. While no one knows what the future of RAM will bring, there will always be advancements in the research and development of all types of computer memory.

Source from https://carcbradios.com/