Memories RAM and ROM
With regard to the hardware of the computers, we understand how the memory the devices that store the data with which the processor works. There are essentially two categories of memories: ROM (Read-Only Memory), which allows only reading data and not loses information in the absence of power; and RAM (Random-Access Memory), which allows the processor to both read as the recording of the data and loses information when there is no electrical power. In this article, the AbbreviationFinder presents the main types of memories ROM and RAM, as well as shows the most important features these devices, such as frequency, latency, tunneling, technology, etc.
The memories ROM (Read-Only Memory – Memory Read-only) receive this name because the data are recorded in them only once. After that, this information is not can be erased or changed, only read by the computer, except by means of special procedures. Another feature of the memories ROM is that they are non-volatile, that is, the recorded data are not lost in the absence electrical power to the device. Here are the main types of memory ROM:
– PROM (Programmable Read-Only Memory): this is a of the first types of ROM memory. The recording data of this type is carried out by means of devices that work through a physical reaction with the elements electric. Once this occurs, the data recorded in the memory The PROM cannot be erased or altered;
– EPROM (Erasable Programmable Read-Only Memory): memories EPROM have as main characteristic the ability to allow that data is rewritten on the device. This is done with the aid a component that emits ultraviolet light. In this process, the data written need be erased completely. Only after that is that a new recording can be made;
– EEPROM (Electrically-Erasable Programmable Read-Only Memory): this type of ROM memory also lets you rewrite data, however, contrary to what happens with the memories EPROM, the processes for erasing and writing data are made electrically, doing that is not necessary to move the device your place for a special apparatus for the rewrite from occurring;
– EAROM (Electrically-Alterable Programmable Read-Only Memory): the memories of EAROM can be seen as a type of EEPROM. Your the main feature is the fact that the data is written to can be changed gradually, which is why this type is generally used in applications that only require rewriting partial information;
– Flash: the Flash memories can also be seen as a type of EEPROM, however, the process of recording (and rewrite) is much faster. In addition addition, Flash memories are more durable and can save a high volume of data. It is possible to know more about this type of memory on the article Cards Flash memory, published here in AbbreviationFinder;
– CD-ROM, DVD-ROM , and the like: this is a category optical disc where the data is written only to a time, to be factory, as the music CDs, or data own user when the own performs the recording. There is also a category that can be compared to the EEPROM, it allows rewriting of data: CD-RW and DVD-RW and the like.
The memories RAM (Random-Access Memory – Memory Random Access) constitute one of the most important parts of computers, since it is therein that the processor stores the the data with which it is dealing. This type of memory has a process of recording data extremely fast, if compared to the various types of ROM memory. However, the recorded information is lost when there is no more electric power, that is, when the computer is off, being, therefore, a type of memory volatile.
There are two types of memory technology RAM that are many used: static and dynamic, that is, SRAM and DRAM, respectively. There is also a newer type called MRAM. Here is a brief explanation of each type:
– SRAM (Static Random-Access Memory – Static RAM): this type is much faster than the DRAM memories, but it stores less data and has high price if we consider the cost per megabyte. Memories SRAM are often used as cache (learn more about cache in this article about processors);
– DRAM (Dynamic Random-Access Memory – Dynamic RAM): memories of this type have high capacity, that is, can hold large amounts of data. However, access to this information tends to be slower than the access to the memories still. This type also tends to have a price much less when compared the static type;
– MRAM (Magnetoresistive Random-Access Memory – RAM Magneto-resistive): the memory MRAM has been studied for a long time, but only in the last few years is that the first units have emerged. It is a type of memory up to a certain point similar to the DRAM but which uses cells magnetic. Thanks to this, these memories consume less energy, are faster and store data for a long time, even in the absence of electrical energy. The problem of the memories, MRAM is they store small amounts of data and are very expensive, therefore, little likely to be adopted on a large scale.
Aspects of the operation of the memories RAM
The DRAM memories are formed by the chips that it contains a quantity the highest of capacitors and transistors. Basically, a capacitor and a transistor together form a memory cell. The first has the function of storing electric current for a certain time, while the second controls the passage of this current.
If the capacitor is storage current, has a bit 1. If you do not is, has a bit 0. The problem is that the information is maintained for a short period of time, and that there there is loss of data from memory, a component of the controller memory is responsible for the function of refresh (or refresh), which consists of rewriting the content of the cell from time to time. Note that this process is performed thousands of times per second.
The refresh is a solution, but accompanied “done side”: this process increases the energy consumption and, consequently, increases the heat generated. In addition, the speed of memory access ends up being reduced.
The memory SRAM, in turn, is quite different from the DRAM and the main reason for this is the fact that it uses six transistors (or four transistors and two resistors) to form a cell memory. In fact, two transistors are responsible for the control task, while the others remain responsible by the electric storage, that is, by the formation bit.
The advantage of this scheme is that the refresh ends up not being necessary, making the SRAM memory is faster and consume less energy. On the other hand, as its manufacturing it is more complex and requires more components, its cost ends up being extremely high, encarecendo for other construction a computer based only on that type. That is why your most common use is as a cache, for so are required small amounts of memory.
As the DRAM memories are more common, they will be the focus of this text from this point on.
CAS and RAS
The processor stores into the RAM the information with which it works, therefore, all of the time, operations recording, deletion, and access to the data are carried out. All of this is possible thanks to the work of a circuit, already referred to, called controller memory.
To facilitate the carrying out of these operations, the memory cells are arranged in a kind array, that is, are oriented in a scheme that recalls lines and columns. The crossing of a certain line (also called a wordline), with a particular column (also called bitline) way what we know as memory address. Thus, to access the address of a position in the memory, the controller gets the value of the column, that is, the value of RAS (Row Address Strobe) and your line value, that is, the value CAS (Column Address Strobe).
Timing and latency of the memories
The parameters of timing and latency indicate how long the memory controller spends with the operations of reading and writing. In general, the smaller this values, the more rapid are the operations.
So that you can understand, take the example of a module memory that informs the following amounts in respect of latency: 5-4-4-15-1T. This value is written in this order: tCL-tRCD-tRP-tRAS-CR. Let’s look at what each of these parameters means:
– tCL (CAS Latency): when an operation memory read is initiated, the signals are triggered to activate the lines (RAS) and the columns (RAS) matching, determine if the operation is read or write (CS – Chip Select), and so on. The parameter CAS Latency indicates, in clock cycles (learn more about clock on this matter about processors), which is the period that there is between the sending of the signal CAS and the provision of the respective data. In other words, it is the interval between the request of a given by the processor and the delivery of this by memory. Thus, in the case of our example, this value is 5 clock cycles;
– tRCD (RAS to CAS Delay): this parameter indicates, also, in clock cycles, the interval between the activation of the line and column of a given site. In the example above, this value corresponds to 4;
– tRP (RAS Precharge): interval in clocks that informs the time between disabling the access to a line and enable access to another. In our example, this value is 4 cycles;
– tRAS (Active to Precharge Delay): this parameter indicates the interval, also in clocks, required between a command activate line and the next action of the same type. In our example, this value is 15 clock cycles;
– CR (Command Rate): the interval between the activation signal CS, and any other command. In general, this value is 1 or 2 clock cycles and is accompanied by the letter T. In our example this value is 1 cycle.
These parameters tend to be informed by the manufacturer in a the label affixed to the comb of memory (many times the value of CMD is not informed). When this does not occur, it is possible to obtain this information through specific software (such as the free CPU-Z for Windows, shown below), or even by the BIOS setup.
The timing parameters to provide a good notion the access time of the memories. Note that, when we talk about addition, we refer to the time that memory takes to provide the data required. What was not said above is that this time it is measured in nanoseconds (ns), that is, 1 second divided by 1,000,000,000.
So, to get a sense of what is the frequency maximum used memory, just divide 1000 by your access time in nanoseconds (this information can be included on a label on the module, or can be reported through special software). For example: if a memory stick works with 15 ns, its frequency is 66 MHz, therefore 1000/15=66.
Some motherboards current or targeted to the public that does overclocking (in a few words, a practice where devices hardware are adjusted so that they work in addition to the specificationsfactory) or, still, the software that detail the characteristics the hardware of the computer, often to inform other parameters, in addition to those mentioned above. Generally, these parameters the additional you are informed as follows: tRC-tRFC-tRRD-tWR-tWTR-tRTP (for example: 22-51-3-6-3-3), also considering clock cycles. Let’s look at what each one means:
– tRC (Row Cycle): consists of the time required to complete a cycle to access a row of the memory;
– tRFC (Row Refresh Cycle): consists of the time required for the execution of cycles of the refresh of the memory;
– tRRD (Row To Row Delay): similar to tRP, but considers the time that the controller needs to wait after a new line has been activated;
– tWR (Write Recovery): tells you the time necessary to to which the memory controller begins to perform an operation writing after performing an operation of the same type;
– tWTR (Write to Read Delay): consists of the time required to which the memory controller begins to execute operations read after you perform a write operation;
– tRTP (Read to Precharge Delay): indicates the time required between a read operation carried out and activation the next signal.
In comparison with other items of a computer, the memories are one of the components that are less energy-consuming. The interesting is that this consumption has decreased with the evolution of technology. For example, modules of DDR2 memory (a technology that still will be discussed in this text), in general, require between 1.8 V and 2.5 V. it Is possible to find combs of DDR3 memory (default that will also be covered in this article) whose requirement is 1.5 V. memory Modules from the old required about 5 V.
Some people with enough knowledge of the subject matter make overclock in the memories by increasing the voltage. With this adjustment, when within certain limits, it is possible to obtain larger levels clock.
SPD (Serial Presence Detect)
The SPD is a small chip (typically of EEPROM) that is entered in the memory modules that contains various information about the specifications of the device, such as the type (DDR, DDR2, etc.), voltage, timing/latency, manufacturer, serial number, etc.
Many motherboards have a setup of BIOS that allows a series of configuration settings. In these cases, a user experience can define the parameters of memory, however, those who do not want to have this work, you can keep the configuration default. Sometimes, this setting is indicated by something related to DPS, as shown in the image below: Detection of errors.
Some mechanisms have been developed to assist in the detection errors of memory, failures, and these can have several causes. These features are especially useful in applications high reliability, such as servers mission critical, for example.
One of these mechanisms is the parity, can only help to detect errors but not correct them. In this scheme, a bit is added to each byte of memory (remember: 1 byte corresponds to 8 bits). This bit takes the value 1 if the amount bit 1 of the byte is even, and assumes the value 0 (zero) if the said amount by odd (the opposite may also happen: 1 for odd and 0 for even). When the reading of data is made, a circuit checks to see if the parity matches the amount of bits 1 (or 0) of the byte. If it is different, an error has been detected.
The parity, however, may not be as precise, since it an error in two bits, for example, can cause the parity bit corresponds to the quantity, odd or even 1-bit of the byte. Thus, for applications that require high precision of the data, you can count on with memories that have ECC (Error Checking and Correction), a mechanism more complex able to detect and correct bit error.
Tunnel types of memory
The tunnel corresponding to the artifact that gives physical form to the memory chips. Here is a brief description of the tunnel types most used by the industry:
– DIP (Dual In-line Package): one of the first types of the tunnel used in memories, being especially popular in the times of the XT and 286. How has terminals contact – “legs” – of great thickness, fit oreven your collage through the weld in the plates can be made easily of way manual;
– SOJ (Small Outline J-Lead): this tunnel receives this name because their terminals contact remember the letter ‘J’. It was quite used in the SIMM modules (seen below), and its shape fixing the plates is made through the weld, not requiring holes in the surface of the device;
– TSOP (Thin Small Outline Package): a type of tunnel whose thickness is considerably reduced in relation to the standards cited earlier (about 1/3 less than the SOJ). By to realize this, terminals of the contact are smaller, in addition to of thinner, decreasing the incidence of interference in the communication. It is a type applied in modules memory SDRAM and DDR (which will be discussed below). There are a variation of this tunnel called STSOP (Shrink Thin Small Outline Package) which is even thinner;
We understand how the module or, even, comb, a small board where they are installed to the tunnels of memory. This plate is attached to the mother board by means of fittings (slots) that are specific to it. Here is a brief description of the most common types of modules:
– SIPP (Single In-Line Pins Package): it is one of the first types of modules that you have come to the market. Is format by chips with encapsulation DIP. In general, these modules were soldiers on the motherboard;
– SIMM (Single In-Line Memory Module): modules of this type were not soldiers, but seated in the motherboard. The first version contained 30 terminal contact (SIMM 30 vias) and was formed by a set of 8 chips (or 9, for parity). With this, they could to transfer one byte per clock cycle. Later came a version with 72 pin (SIMM of 72 lanes), therefore, larger and able to download 32-bit at a time. Modules SIMM 30 vias could be found with capacities ranging from 1 MB to 16 MB. The SIMM modules (72-way, for its time, were commonly found with capacities ranging from 4 MB 64 MB;
– DIMM (Double In-Line Memory Module): modules DIMM lead this name for having terminal contacts on both sides comb. Are able to transmit 64 bits at a time. The first version – applied in memory SDR SDRAM – had 168 pins. Then they were released modules 184-way, used memories DDR, modules, 240-way, used in modules DDR2 and DDR3. There is a standard DIMM of reduced size called SODIMM (Small Outline DIMM), which are used mainly on portable computers, like notebooks;
– RIMM (Rambus In-Line Memory Module) is formed by 168 way, this module is used for the memories Rambus, that will be addressed later in this article. A curious fact is that for each memory stick Rambus installed on the computer is need to install a module “empty”, 184 way, called C-RIMM(Continuity-RIMM).
Technologies of memories
Various technologies of memories were (and are) created with the passing of time. It is thanks to this that, periodically, we find faster memories with higher capacity and even memories that require increasingly less energy. Here is a brief description of the main types of RAM:
– FPM (Fast-Page Mode): one of the first technologies the RAM memory. With the FPM, the first reading of the memory has an access time that is greater than the readings of the following. This is because there are made, in fact, four read operations in succession, instead of only one, in a scheme of the type x-y-y-y, for example: 3-2-2-2 or 6-3-3-3. The first reading ends up being more time consuming, but the following three are the fastest. This is because the memory controller works only once with the address of a line (RAS), and then work with a sequence of four columns (CAS), instead of working with a signal RAS and a CAS for each bit. Memories FPM used modules SIMM, both 30 and 72 way;
– EDO (Extended Data Output): the successor of the technology FPM is the EDO, which has as a highlight the ability to allow that a memory address is accessed at the same time in which a previous request is still in progress. This type has been mainly applied in the SIMM modules, but also came to be found in modules, DIMM 168 way. There was also a similar technology, called BEDO (Burst EDO), which worked more quickly by having access time less, but almost not was used, since it had a higher cost because of the property of the company Micron. In addition, it was “overshadowed” by the arrival of the technology SDRAM;
– SDRAM (Synchronous Dynamic Random Access Memory): the memories FPM and EDO are asynchronous, which means that does not work synchronously with the processor. The problem is that, with processors that are each faster and faster, it began to become a problem, because many times the processor had to wait too much to have access to the data in the memory. The memories SDRAM, in turn, are synchronized with the processor, avoiding the problems of delay. From this technology, it is now considering the frequency with which the memory works to a measure of speed. Arose then the memories the SDR SDRAM (Single Data Rate SDRAM), which could work with 66 MHz, 100 MHz and 133 MHz (called PC66, PC100 and PC133 sdram, respectively). Many people refer to this memory simply as “memories SDRAM” or even as “memories DIMM”, because of your module. However, the name SDR is the most appropriate;
– DDR SDRAM (Double Data Rate SDRAM): memories DDR feature significant progress in relation to the standard SDR, this is because they are able to cope with the double data in each clock cycle (memoirs SDR only work with an operation cycle-by-cycle). Thus, a DDR memory working frequencyof 100 MHz, for example, ends up doubling your performance, as if he worked at the rate of 200 MHz. Visually, it is possible to identify them easily in relation to the modules in the SDR, because thisthe latter contains two divisions at the bottom, where are your contacts, while the memories DDR2 have only a division. You can learn more about this technology in the matter of Memory DDR, published here in AbbreviationFinder;
– DDR2 SDRAM: as the name indicates, the memories DDR2 are an evolution of the memories DDR. Its main feature it is the ability to work with the four operations per clock cycle, therefore, twice the previous standard. The modules DDR2 also have only a division in its lower part, however, this openness is a little more shifted to the side. Learn more about this technology in the field MemoryDDR2, released here in the AbbreviationFinder;
– DDR3 SDRAM: the DDR3 memory are, of course, an evolution of the memories of DDR2. Again, here the fold-if the amount of operations per clock cycle, this time, eight. A novelty here is the possibility of the use of Triple-Channel. Learn more about this type in this article on DDR3;
– Rambus (Rambus DRAM): the memories Rambus receive this name for being a creation of the company Rambus Inc. and came on the market with the support of Intel. They are different from standard SDRAM, since it only work with 16 bits at a time. In compensation, memories Rambus working with a frequency of 400 MHz and with two operations per clock cycle. Had as disadvantages,however, rates of latency very high, heating and high a greater cost. Memories Rambus never had much acceptance in the market, but also have not been a total fiasco: were used, for example, in the game console Nintendo 64. Interestingly, the memories Rambus work in pairs with “modules empty” or “combs blind”. This means that, for each module Rambus installed, an “empty module” must to be installed in another slot. This technology ended up losing space for the memories DDR.
With the passing of time, the evolution of the technologies of memories not only makes them faster, but also makes with that they count on a larger data storage capacity. Memories ROM type Flash, for example, can store multiple gigabytes. With regard to the memories RAM, the same occurs. On account of this, the natural question is: how much to use? The answer depends on a number of factors, however, the industry not to work to further increase the speed and the capacity of these devices. So, do not be afraid: when unless you wait, you will hear talk of a new memory technology that may become a new pattern of the market 🙂