A, a ROM (the Read Only Memory)
a ROM (the Read Only Memory), a read only memory. To store and preserve data. ROM data can not be free update, but can be read at any time. Even the power outage, ROM data can be retained.
There are many ROM:
a PROM is a one-time programmable (not modify) a ROM;
an EPROM UV erasable programmable ROM;
EEPROM is an electrically erasable programmable ROM, delete and rewrite byte, write time is long, the writing is slow; now widely used as nonvolatile data storage. It characterized by random access and modify any of the bytes to be written into each bit of 0 or 1. This is the most traditional kind of EEPROM, data after power is not lost, you can save 100 years, you can erase 100w times. High reliability, but the circuit complexity / cost is high. Therefore, the current EEPROM are tens of kilobytes to a few hundred kilobytes, rarely have more than 512K of.
Two, the RAM (the Random Access Memory)
the RAM (the Random Access Memory), a random access memory. Is a direct exchange with the CPU internal data memory, also called the memory. It can be read at any time, and fast, usually as the operating system or other programs that are running in the temporary data storage medium, when the power is turned off the data RAM can not be retained. RAM can be further divided into static RAM (SRAM) and dynamic memory (DRAM) two categories.
Static RAM (Static RAM / SRAM): SRAM is very fast, does not require refresh circuitry that is able to save the data, is read fastest storage device, but the lower the degree of integration, very expensive, multi-level cache for the CPU , the secondary cache (L1 / L2 cache).
Dynamic RAM (Dynamic RAM / DRAM), DRAM data retention time is very short (requires memory refresh circuit, every once in a while, refresh charging time, otherwise the data will disappear), the speed is slower than SRAM, but it is still better than any of ROM faster, but the price is much cheaper DRAM compared to SRAM, DRAM computer memory is of.
DRAM is divided into many common main FPRAM / FastPage, EDORAM, SDRAM, DDR RAM, RDRAM, SGRAM and WRAM and so on, which introduces a DDR RAM here.
DDR RAM (Date-Rate RAM) also known as the DDR SDRAM, a RAM and such modified SDRAM are substantially the same, except that it can read and write data twice in one clock, so that data transfer speed is doubled . This is currently the most frequently used computer memory, and it has a cost advantage, Intel actually beat another kind of memory standard -Rambus DRAM. In many high-end graphics card, also equipped with high-speed DDR RAM to increase bandwidth, which can greatly improve the pixel rendering capabilities of 3D accelerator card.
Memory works: (ie running) data and program memory is used to store currently being used, we usually mentioned computer memory refers to dynamic memory (ie, DRAM), dynamic memory in the so-called "dynamic" means that when we write data to DRAM, after a period of time, data will be lost, and therefore need an additional reset circuit for memory refresh operation.
Specific work process is as follows: the storage unit stores a DRAM is 0 or 1 depending on whether there is a charge capacitor, the charge represents 1, represents 0 uncharged. But over time, the capacitor will discharge represents 1, 0 represents the capacitance will absorb charges, which is cause of data loss; refresh operation periodically checks the capacitance, if the power is greater than 1/2 of the full amount, that its representatives 1 and the capacitor is fully charged; if the level is less than 1/2, it is considered representative of 0, and the capacitor is discharged, thereby to maintain the continuity of the data.
Three, FLASH Memory
(. 1) FLASH memory
FLASH memory, which belongs to a memory device, a non-volatile (Non-Volatile) memory, combines the advantages of ROM and RAM, not only an electronic erasable programmable ( EEPROM) performance, power outage will not lose data and can quickly read data (NVRAM advantage), U disk and MP3 use is in this memory. Over the last 20 years, embedded systems have been using ROM (EPROM) as their storage device, but in recent years instead of Flash comprehensive position in the embedded system ROM (EPROM), used as a storage and operating system or program Bootloader direct use code or when the hard disk (U disk). The biggest difference in the broad sense FLASH ROM belongs, EEPROM and FLASH by sector operation is improved with respect to the EEPROM is not erased when the bytes but in units of blocks, a simplified circuit, data density higher, reducing the cost. M on the ROM are generally FLASH. The operation is byte EEPROM. Currently there are two main Flash and NOR Flash NADN Flash.
NOR Flash: NOR Flash read and read our common SDRAM is the same, the user can run directly loaded NOR FLASH inside the code, thus reducing the capacity of SRAM and thus cost savings. Generally small capacity with NOR Flash, because of its fast read speed, multi-operating system to store important information and the like. NOR FLASH separate data and address lines, may be implemented as a RAM random addressing can be read any byte, but still erased blocks Yaoan wipe.
NADN Flash: random access memory not taken technology, it is read in the form of one of the first reading is performed, usually a 512 byte read, using this technology Flash relatively inexpensive. Users can not directly run the code on NAND Flash, so a lot of use of NAND Flash development board in addition to the use of NAND Flah, it also made a small piece of NOR Flash to run the boot code. With large-capacity NAND FLASH, NAND FLASH most common application is an embedded system using the DOC (Disk On Chip) and we usually use "flash", you can erase online. NAND FLASH same is erased in blocks, the data and address lines multiplexed address lines can not use random addressing. Can only be read by page read. (NAND FLASH erased in blocks, read page by page, NOR FLASH no page). Since the NAND FLASH pin multiplexing, thus reading speed slower than that NOR FLASH, erase and write but much faster than NOR FLASH. NAND FLASH internal circuit simpler, so large data density, small size, low cost. So are large-capacity flash NAND FLASH type. 2 ~ 12M is mostly small-capacity flash NOR FLASH type.
FLASH currently on the market mainly from Intel, AMD, Fujitsu and Toshiba, while NAND Flash production of major manufacturers are Samsung and Toshiba. According to different manufacturers and different applications, flash memory cards about U disk, SmartMedia (SM Card), Compact Flash (CF card), MultiMediaCard (MMC cards), Secure Digital (SD card), Memory Stick (Memory Stick), MS card, TF card (micro SD card), PCIe flash card, XD-Picture card (XD card) and Microdrive (MICRODRIVE) flash memory cards, although these different appearance, specifications, but the principle is the same technology.
FLASH works: Flash Memory, belonging to a nonvolatile storage device (Non-volatile Memory Device), Flash memory is internal MOSFET, which has a suspended gate (Floating Gate), is a unit of real data storage. Data is a charge (electrical charge) stored in the Flash memory cell. Number, depending on the outer sector of FIG (external gate) voltage is applied to the stored charge, which controls the charge is flushed to the storage unit or to release charges. The data indicates, whether the voltage to the charge stored exceeds a specific threshold value Vth is represented.
For representing data, a single memory cell in the charge stored in the internal voltage, and a certain threshold voltage Vth, as compared, if Vth is larger than this value, 1 is represented, on the contrary, is less than Vth, it represents 0; for nand Flash 1 data is written, is to control External Gate charge, enough so that the stored charge, exceeds the threshold voltage Vth, it means 1 a. For writing 0, the discharge is reduced to less than Vth of the charge, it indicates the 0
(2) SSD and an HDD
an HDD: hard disk drive (Hard Disk Drive), basic computer memory, i.e. mechanical hard drives, store information on the magnetic field, it is also called disk. Here do not do too much introduction.
SSD: SSD (Solid State Drives), referred to the solid disk. By a control unit and a storage unit (FLASH chips, DRAM chips) composition. SSD storage medium is divided into two types, one is to use a flash memory (FLASH chip) as a storage medium, another is the use of DRAM as a storage medium.
DRAM-based categories: using DRAM as a storage medium, narrow range of applications. It is designed to emulate traditional hard drive, it can be carried out volumes set up and manage most of the operating system's file system tools, and provides industry-standard PCI and FC interface to a host or server. Application of SSD can be divided into two kinds of hard and the SSD array. It is a high performance memory, and a very long life, independent power drawback is the need to protect the data. DRAM SSDs are relatively non-mainstream device.
Flash-based categories: SSD uses flash-based FLASH chip as a storage medium, which is also known as the SSD. Its appearance can be made notebook drives, micro drives, memory cards, U-style. An SSD within the body is actually a PCB, the PCB and the piece parts that control basic chip, the cache chip (the lower end portion of the hard disk without cache chips), and flash memory chips for storing data.
Cache chip: next to the main chip is a chip cache, like the need to SSDs and traditional hard disk high-speed cache chip secondary master chip for data processing. It should be noted that there are some low-cost solid-state drive solutions in order to save costs, eliminating the need for this cache chip, so there will be some impact on performance when using.
Flash memory chips: In addition to the main chip and cache chip, PCB board most of the rest position of the flash memory chips are NAND Flash. NAND Flash flash memory chip is divided into the SLC (Single Level Cell) the MLC (multi level cell) and TLC (triple cell) NAND flash memory.
The difference between SLC, MLC, TLC NAND are
for devices based on NAND memory technology, whether it is U disk or SSD, or even SD card, will involve a question of cost, so the product design transition from SLC to MLC, and then to TLC, even QLC will come out in the follow-up, then the SLC, MLC, TLC what is the impact on users?
SLC - full name SLC (Single Level Cell - SLC) i.e. Single Level Cell
SLC technology is characterized by a thinner oxide film and electrode among source, when data is written on the floating gate through the floating gate the charge-summing voltage, then the electrode can eliminate the stored charge through the source, in such a way, one can store information unit, this technique provides fast programming and reading, but by this technique limited Silicon efficiency issue that must be strengthened by the more advanced process technology (process enhancements), in order to enhance upward SLC process technology.
MLC - MLC English name (Multi Level Cell - MLC) that is multi-layered storage
Intel (Intel) in September 1997 the first successful development of MLC, whose role is to place information from two units into a Floating Gate (Flash part of the charge stored in the storage unit), and then using the different potentials (Level) of the charge, the voltage stored in the memory write control precision. MLC by using a large number of voltage levels, each cell stores two bits of data, the data density is relatively large. SLC structure two values 0 and 1, while the MLC architecture can store more than one value 4, therefore, MLC architecture can have better storage density.
TLC - TLC full name (Trinary-Level Cell) i.e., three-layer reservoir
TLC i.e. 3bit per cell, each cell can store multiple data MLC than 1/2, a total of eight charge value, required a longer access time, Thus the transmission speed is slower. TLC advantage of cheap production cost per megabyte is the lowest, but life is short, only about 1,000 Endurance.
As described above, TLC and then, the precise Cell higher control voltage from SLC to MLC, which led directly TLC life dropped to only 1000 PE, corresponding to the SLC and MLC respectively 3000 and 10,000, relatively speaking TLC durability significantly.
TLC Another disadvantage is the efficiency of reading and writing data in the SLC age, a time cell only needs to read / write a 'bit, to the times per MLC needs to read / write 2bit, but rose to TLC Age to 3bit, it is clear that its performance by voltage control the complexity of the program will slow down, of course, due process and master continually upgraded and now TLC already tied MLC products.
But a short time TLC durable flawed and can not be effectively addressed, of course, TLC durable wear can be balanced by increasing the capacity of the storage device, disguised extend the life of the product
(3) eMMC and USF2.0
in recent years the rise of mobile phones, mobile phone flash memory standard has been greatly improved, eMMC specification standards gradually from eMMC 4.3 era to today's eMMC 5.0 storage products, cell phone flash performance With a very large increase. But now a new flash pass specifications have emerged, it is the UFS 2.0 flash transfer standard, has read faster performance than the eMMC 5.0.
eMMC flash pass specifications: eMMC is called the "embedded Multi Media Card", by the MMC Association entered into primarily for embedded memory standard phone or tablet and other products. eMMC is a clear advantage in the package of an integrated controller, which provides a standard interface and a flash memory management, it utilizes the eMMC main controller, a flash memory integrated into particles within a small BGA package.
eMMC = NAND Flash + controller + standard package interface
now, eMMC 4.5 has been around, eMMC 4.4 read speeds of about 104MB / s, eMMC 4.5 compared to 200MB / s, the performance was also very good; and in 2013 7 May 29 Samsung eMMC 5.0 storage industry's first mass production of products, its read speed of 400MB / s, but because of the use of 8-bit parallel interface, so the performance potential has substantially reached a bottleneck, with the latest eMMC 5.1 specification for , the theoretical bandwidth of 600MB / s or so, performance boost is substantially impossible.
UFS 2.0: UFS 2.0 flash memory specification is adopted a new standard that uses a serial interface, much like PATA, SATA conversion. And it supports full-duplex operation, can read and write operations are also supported instruction queue. In contrast, the eMMC is half-duplex, the reader must be performed separately, instructions are packaged, the speed is already in a slightly inferior. And not only UFS chip transmission speed, lower power consumption than half eMMC 5.0, it can be said to be the ideal flagship phone with flash.
In fact, JEDEC in September 2013 has released a new generation of flash storage standard UFS 2.0, UFS 2.0 flash memory can read and write speeds of up to 1400MB per second, which is equivalent to two read-write CD-ROM within two seconds of data, not only have more than eMMC huge advantage, and it even allows the use of solid state disk flash memory storage media on the computer also dwarfs.
In addition to speed has a huge advantage in terms of performance than in terms of power consumption UFS 2.0 also has better performance. In fact, if you compare the power consumption, even though a new generation of UFS 2.0 standard can also flat with eMMC. However eMMC 2.0 and UFS power consumed during operation is about 1mW (mW), and the power consumption in standby mode will be lower than 0.5mW. UFS 2.0 when fully loaded, the power actually consumed more than the eMMC, but it can be done faster and earlier operation is switched to the standby state, the performance in terms of power consumption and UFS 2.0 eMMC comparable.
In the transmission speed, UFS 2.0 is far more than can be said of eMMC. Even compared with the current latest eMMC 5.0 standard, UFS 2.0's speed three times higher. UFS 2.0 full-duplex mode can be exchanged between the host and the flash memory UFS, write may be performed simultaneously. Further, UFS 2.0 additional control channels may effectively ensure the secure transmission of data, read and write operations do not have to do because unnecessary waiting, which is the key UFS 2.0 higher speed. It is understood, UFS 2.0 there are two versions, has two transmission channels. Theoretical bandwidth of HS-G2 have 5.8Gbps, which is more than 740MB / s, HS-G3 is doubled to 11.6Gbps, close to 1.5GB / s, speed UFS 2.0 victory.
Summary: While not as eMMC now in the latest UFS 2.0 transfer speeds and performance, but in the future for a long period of time, or will it continues to occupy the mainstream of mobile products of flash memory chips. Because eMMC has a more mature technology, capable of mass production, lower cost and more suitable for popular mobile products flash memory chip demand; and UFS 2.0 was beautiful, but the higher cost of production, beginning in their early market now, can only It is placed on the use of high-end products. But now the trend, UFS 2.0 will gradually become the mainstream mobile market is believed to be no problem, after all, is the need for progress in science and technology.
Fourth, added that
what is SDRAM?
SDRAM (Synchronous DRAM): synchronous dynamic random access memory. The current bandwidth of 64bit line memory 168 basically using SDRAM chips, 3.3V operation voltage, high access speed 7.5ns, the EDO memory fastest 15ns. The CPU and the RAM at the same clock frequency controls the RAM and CPU FSB synchronization, cancel the waiting time. Therefore, the transmission rate faster than EDO DRAM.
What is SDRAM DDR?
DDR (Double the Data Rate) SDRAM. Its core is based on SDRAM, but the speed has improved. SDRAM read only data on the rising edge of the clock signal, and the DDR data are read in the rising and falling edges of the clock signal, therefore, it is 2 times the standard speed of the SDRAM.
What is the RDRAM?
RDRAM (Rambus DRAM): Bus dynamic random access memory, it is a company with a patented technology RAMBUS INTEL companies proposed, its data transfer rate of up to 800MHZ, and it was only 16bit bus width , far less than the 64bit SDRAM now.
What is SPD?
SPD (Serial Presence Detect): SPD is an 8-pin EEPROM (Electrically Erasable Programmable ROM electrically EPROM) with a capacity of 256 bytes, which is mainly relevant information stored in the memory such as capacity, chip manufacturers, memory module manufacturers and working speed. SPD contents are generally written by the memory module manufacturers. SPD motherboard supports automatic detection of data in the SPD at startup, and thus set the operating parameters of memory.
Now microcontroller, RAM is doing the runtime data memory, FLASH is the main program memory, EEPROM primarily need to save some of the non-volatile data in the program is running. In addition, a number of variables, are placed in the RAM, some initialization data such as the contents of the LCD screen to be displayed, are placed in the district of FLASH (that is to say before the district ROM), EEPROM may or may not, exist primarily in the operation of some data after power failure without loss.
