FPM memory
FPM is the abbreviation of fast page mode, which is commonly used in early personal computers. It transmits data every three clock cycles. Has been eliminated.
Extended data output
EDO is the abbreviation of extended data output. It cancels the time interval between two storage cycles of motherboard and memory, and transmits data every two clock cycles, which greatly shortens the access time and improves the access speed by 30% to 60ns. EDO memory is mainly used for 72-wire SIMM memory chips and PCI graphics cards with EDO memory chips. This kind of memory is very popular in 486 and early Pentium computer systems. Divided into 72 lines and 168 lines. Its working voltage is 5V and its bandwidth is 32 bits. Two or four rows must be used in pairs. It can be used on the motherboard of Intel 430FX/430VX or even 430TX chipset. At present, it has also been eliminated and can only be seen on some old machines.
Synchronous dynamic random access memory
SDRAM is the abbreviation of synchronous dynamic random access memory, and it is a widely used memory form in previous years. SDRAM uses a working voltage of 3.3v and a bandwidth of 64 bits. SDRAM locks CPU and RAM together through the same clock, so that RAM and CPU can enjoy a clock cycle and work synchronously at the same speed, which can be 50% faster than EDO memory. SDRAM is based on dual bank structure, which includes two staggered memory arrays. When the CPU accesses data from one memory bank or array, the other is ready to read and write data. By tightly switching these two memory arrays, the reading efficiency can be doubled. SDRAM is not only used as main memory, but also widely used in video memory on graphics cards. SDRAM used to be the mainstream memory for a long time, and it was supported from 430TX chipset to 845 chipset. But with the popularization of DDR SDRAM, SDRAM is gradually withdrawing from the mainstream market.
RDRAM memory
RDRAM is the abbreviation of Rambus dynamic random access memory, which is a memory with system bandwidth and inter-chip interface design developed by Rambus company. It can transmit data through a simple bus in a very high frequency range, and at the same time use low-voltage signals to transmit data at two edges of high-speed synchronous clock pulses. At first, the Intel 820 chipset supported RDRAM, and then there were 840,850 chipsets and so on. RDRAM was initially strongly supported by Intel, but due to its high price and patent licensing restrictions of Rambus, it has not been able to become the mainstream of the market, and its position was quickly replaced by DDR SDRAM, which is relatively cheap and has the same excellent performance, with a small market share.
DDR SDRAM memory
Definition of DDR2:
DDR2 (Double Data Rate 2) SDRAM is a new generation of memory technology standard developed by JEDEC (Joint Committee on Electronic Equipment Engineering). The biggest difference between DDR2 and the previous generation memory technology standard is that although the basic mode of data transmission is adopted while the clock rises/falls, the pre-reading ability of DDR 2 memory is twice that of the previous generation DDR memory (that is, 4-4bit data reading and prefetching). In other words, DDR2 memory can read/write data at 4 times the speed of external bus and run at 4 times the speed of internal control bus.
The difference between DDR2 and DDR:
1, delay problem:
As can be seen from the above table, under the same core frequency, the actual working frequency of DDR2 is twice that of DDR. This is because the 4-4 bit read-ahead capability of DDR2 memory is twice that of standard DDR memory. In other words, although DDR2, like DDR, adopts the basic method of transmitting data at the same time with rising and falling clock delays, DDR2 has twice the ability of pre-reading system command data. That is to say, under the same working frequency 100MHz, the actual frequency of DDR is 200MHz, while DDR2 can reach 400MHz.
Then there is another problem: in DDR and DDR2 memories with the same working frequency, the storage delay of the latter is slower than that of the former. For example, DDR 200 and DDR2-400 have the same delay, while the latter has twice the bandwidth. In fact, the bandwidth of DDR2-400 and DDR400 is the same, which is 3.2GB/s, but the core working frequency of DDR400 is 200MHz, while the core working frequency of DDR2-400 is 100MHz, which means that the delay of DDR2-400 is higher than that of DDR 400.
2, packaging and calorific value:
The biggest breakthrough of DDR2 memory technology lies not in the transmission capacity that users think is twice that of DDR, but in the case of lower calorific value and power consumption, DDR2 can achieve faster frequency increase and break through the 400MHZ limit of standard DDR.
DDR memory is usually packaged with TSOP chip, which can work well at 200MHz. When the frequency is high, its long pin will produce high impedance and parasitic capacitance, which will affect its stability and the difficulty of frequency improvement. This is also the reason why DDR core frequency is difficult to break through 275MHZ. DDR2 memory is packaged in FBGA. Different from the widely used TSOP package at present, FBGA package provides better electrical performance and heat dissipation, which provides a good guarantee for the stable operation and future frequency development of DDR2 memory.
The voltage of DDR2 memory is 1.8V, which is much lower than the DDR standard of 2.5V, thus providing significantly less power consumption and heat generation. This change is of great significance.
New technologies adopted by DDR2:
In addition to the above differences, DDR2 also introduced three new technologies, namely OCD, ODT and Post CAS.
OCD (Off-chip Drive): The so-called off-line drive adjustment, DDR II can improve the signal integrity through OCD. DDR II equalizes the two voltages by adjusting the pull-up resistance. OCD is used to improve signal integrity by reducing the inclination angle of DQ-DQS; Improve signal quality by controlling voltage.
ODT:ODT is the terminal resistance of the built-in core. We know that the motherboard using DDR SDRAM needs a lot of termination resistors to prevent the data line from reflecting signals. Greatly increasing the manufacturing cost of the motherboard. In fact, different memory modules have different requirements for termination circuits. The size of the terminal resistance determines the signal ratio and reflectivity of the data line. If the termination resistance is small, the signal reflection of the data line is low, but the signal-to-noise ratio is also low. When the termination resistance is high, the signal-to-noise ratio of the data line is high, but the signal reflection will also increase. Therefore, the termination resistance on the motherboard can't match the memory chip well, and it will also affect the signal quality to some extent. DDR2 can construct suitable termination resistors according to its own characteristics to ensure the best signal waveform. Using DDR2 can not only reduce the motherboard cost, but also obtain the best signal quality, which is incomparable to DDR.
Post CAS: designed to improve the utilization efficiency of DDR II memory. In the post-CAS operation, the CAS signal (read/write/command) may be inserted into the clock cycle after the RAS signal, and the CAS command may remain valid after an extra waiting time. The original tRCD(RAS to CAS and delay) is replaced by al (additive delay), which can be set at 0, 1, 2, 3, 4. Since the CAS signal is one clock cycle later than the RAS signal, the ACT and CAS signals will never conflict.
Generally speaking, DDR2 has adopted many new technologies and improved many shortcomings of DDR. Although there are many shortcomings such as high cost and slow performance, I believe that with the continuous improvement and perfection of technology, these problems will eventually be solved.
DDR SDRAM is the abbreviation of double data rate synchronous dynamic random access memory, which is a memory standard proposed by VIA and other companies to compete with RDRAM. DDR SDRAM is an updated product of SDRAM, which uses a working voltage of 2.5v It allows data transmission at the rising and falling edges of clock pulses, so that the speed of SDRAM can be doubled without increasing the clock frequency, and the transmission rate and memory bandwidth are twice that of SDRAM. For example, compared with PC 133 SDRAM, the working frequency is also 133MHz, but memory bandwidth reaches 2. 12 GB/s, which is twice as high as PC 133 SDRAM. At present, mainstream chipsets support DDR SDRAM, which is the most commonly used memory type.
DDR2 memory
ECC is not a memory type. ECC (Error Correction Coding or Error Detection and Correction) is a kind of memory with automatic error correction function. Intel's 82430HX chipset began to support. Motherboards using this chipset can install and use ECC memory. However, due to the high memory cost of ECC, it is mainly used in commercial computers with high requirements for system reliability, such as servers/workstations. In fact, memory errors do not occur frequently, and ordinary motherboards do not support ECC memory, so ordinary home and office computers do not need to use ECC memory.