Memory strip DDR and DDR2 differential Interpretation _ Hardware Maintenance

Dual-channel memory dual-channel memory technology is actually a memory control and management technology, which relies on the memory controller of the chipset to take effect, theoretically able to make two of the same size of memory to provide the bandwidth of one-fold increase. It is not a new technology that has long been applied to servers and workstation systems, but only to the Desktop board technology front desk to address the increasingly embarrassing memory bandwidth bottlenecks in desktops. A few years ago, Intel Corporation had launched a I820 chipset that supported dual-channel memory transfer technology. It and RDRAM memory constitute a pair of gold partner, play out of the outstanding performance so that it has become the market's biggest bright spot, but the production cost is too high the flaw has caused the applause not to draw the case, finally by the market eliminated. Since Intel has abandoned support for RDRAM, the current dual-channel memory technology for mainstream chipsets refers to dual-channel DDR memory technology, the main dual-channel memory platform Intel is the Intel 865, 875 series, and AMD is the Nvidia Nforce2 series.
Dual-channel memory technology is a low-cost, high-performance solution to the problem of CPU bus bandwidth and memory bandwidth. Now that the CPU's FSB (front bus frequency) is getting higher, Intel Pentium 4 has a much higher demand for memory bandwidth than AMD Athlon XP. The Intel Pentium 4 processor and North Bridge chip data transfer using QDR (Quad data Rate, four data transmission) technology, the FSB is 4 times times the FSB. Intel Pentium 4 FSB is 400, 533, 800MHz, bus bandwidth is 3.2gb/sec,4.2gb/sec and 6.4gb/sec respectively, and DDR 266/DDR 333/DDR 400 can provide the memory bandwidth is 2.1gb/ Sec,2.7gb/sec and 3.2gb/sec. In single channel memory mode, the DDR memory does not provide the data bandwidth required by the CPU and thus becomes a performance bottleneck for the system. In dual-channel memory mode, dual-channel DDR 266, DDR 333, DDR 400 can provide the memory bandwidth is 4.2gb/sec,5.4gb/sec and 6.4gb/sec, where you can see that dual-channel DDR 400 memory just can meet the 800MHz FSB Pentium 4 Processor bandwidth requirements. For AMD Athlon XP platform, its processor and North Bridge chip data transmission technology using DDR (double data Rate, double data transmission) technology, FSB is the FSB twice times, its demand for memory bandwidth is much lower than the Intel Pentium 4 platform, Its FSB 266, 333, 400MHz, bus bandwidth is 2.1gb/sec,2.7gb/sec and 3.2gb/sec, the use of single channel DDR 266, DDR 333, DDR 400 to meet its bandwidth requirements, so in AMD The use of dual-channel DDR memory technology on the K7 platform is less effective and performance improvement is not as obvious as the Intel platform, and the most significant performance impact is the integration of integrated motherboards.
The nforce chipset introduced by NVIDIA is the first chipset to extend the DDR memory interface to 128-bit, and Intel also uses this dual-channel DDR memory technology on its E7500 server board chipset, and SIS and via are responding, Actively develop this technology that will multiply the DDR memory bandwidth. However, for a variety of reasons, the transfer of this dual-channel DDR (128 bit parallel memory interface) is no easy feat for many chipset vendors. DDR SDRAM memory is completely different from rdram memory, which has a high latency feature and is serial transmission, which determines the difficulty and cost of designing a dual-channel RDRAM memory chipset. But DDR SDRAM memory has its own limitations, it is a low latency characteristics, the use of parallel transmission mode, but also the most important point: when the DDR SDRAM operating frequency is higher than 400MHz, its signal waveform often appear distortion problem, These are all designed to make a chipset that supports dual-channel DDR memory systems a little more difficult, and the manufacturing cost of the chipset increases correspondingly, and these factors constrain the development of this memory-control technology.
The common single channel memory system has a 64-bit memory controller, while the dual-channel memory system has 2 64-bit memory controller, in dual-channel mode with 128bit of memory bit width, so that the memory bandwidth in theory to increase by one times. Although the bandwidth provided by a dual 64-bit memory system is equivalent to the bandwidth provided by a 128-bit memory system, the results are different. The dual-channel system contains two independent, complementary intelligent memory controllers, and in theory, two memory controllers can operate at the same time with 0 delays between them. For example, two memory controllers, one for a and another for B. When controller B is ready for the next access memory, controller A is reading/writing primary memory, and vice versa. This complementary "nature" of two memory controllers can reduce the wait time by up to 50%. The two memory directors for dual-channel DDR are functionally identical, and the timing parameters of the two controllers can be programmed individually. This flexibility allows the user to use two different structure, capacity, speed DIMM memory, when the dual-channel DDR simply adjusts to the lowest memory standard to achieve 128bit bandwidth, allowing the different density/latency characteristics of the DIMM memory can be reliably common operation.
Desktop chipsets that support dual-channel DDR memory technology, Intel platforms with Intel's 865P, 865G, 865GV, 865PE, 875P, and 915, 925 series, via Pt880,ati's Radeon 9100 IGP series, SIS's Siis 655,sis 655FX and SIS 655TX;AMD platforms have via Kt880,nvidia's nForce2 Ultra 400,nforce2 Igp,nforce2 spp and their subsequent chips.
AMD's 64-bit CPU, because of the integration of memory controller, so whether to support memory dual-channel look at the CPU can. At present, AMD Desktop CPU, only 939 interface support memory dual channel, 754 interface does not support memory dual channel. In addition to AMD's 64-bit CPU, whether other computers can support memory dual channels depends primarily on the motherboard chipset, supports dual-channel chipsets with descriptions, and can view motherboard chipset data. In addition, some chipsets in theory support the different capacity of the memory bar to achieve dual channel, but actually it is recommended to use a consistent two of the parameters of the memory bar.
Memory dual-channel generally requires the memory slot on the motherboard to use the color, in addition to some motherboards also need to do in the BIOS settings, General motherboard specifications will be described. When the system has implemented dual-channel, some motherboards will be prompted at post, you can take a closer look. Since the self-test is faster, it may not be visible. So you can use some software to view, many software can check, such as cpu-z, relatively small. In the term "memory" there is a "channels" project, if the word "Dual" is shown here, it means that two channels have been implemented. Two 256M of RAM makes the dual channel effect better than a 512M memory because one memory cannot form a dual channel.
The difference between DDR and DDR2
Strictly speaking DDR should be called DDR SDRAM, people used to call DDR, some beginners also often see DDR SDRAM, it is considered SDRAM. DDR SDRAM is the abbreviation for Double Data Rate SDRAM, which is the meaning of double speed synchronous dynamic random memory. DDR memory is developed on the basis of SDRAM memory, still follow the SDRAM production system, so for the memory vendors, only a little improvement of the equipment to create common SDRAM, can achieve the production of DDR memory, can effectively reduce costs.
SDRAM transmits only once data in a clock cycle, it is in the clock rising time to carry on the data transmission, but the DDR memory is a clock cycle transmits two times the data, it can transmit the data each time in the clock rise time and the descent period, therefore is called the double rate synchronous dynamic random memory. DDR memory can achieve higher data transfer rates at the same bus frequency as SDRAM.
Compared to SDRAM: DDR uses a more advanced synchronization circuit, so that the specified address, data transmission and output of the main steps are performed independently, but also to maintain full synchronization with the CPU; DDR used DLLs (Delay Locked Loop, delay-locked loop to provide a data filter signal) technology, When the data is valid, the storage controller can use this data filtering signal to pinpoint data, output once every 16 times, and resynchronize data from different memory modules. DDL essentially does not need to raise the clock frequency to increase the SDRAM speed, it allows the clock pulse to rise along and down along the reading data, so its speed is the standard Sdra twice times.
DDR differs from SDRAM in terms of shape volume, and they have the same size and the same pin distance. But the DDR is 184 pins, 16 more pins than SDRAM, including new controls, clocks, power supplies and grounding signals. DDR memory uses the SSTL2 standard that supports the 2.5V voltage, rather than the 3.3V voltage LVTTL standard used by SDRAM.
DDR2 memory start frequency from the DDR memory highest standard frequency 400Mhz, has been defined can be produced by the frequency support to 533Mhz to 667Mhz, the standard frequency of work frequency is 200/266/333mhz, operating voltage of 1.8V. DDR2 is completely incompatible with the 184PIN DIMM interface standard for DDR, using the newly defined PIN DIMM interface standard.
DDR2, like DDR, uses a basic way of transmitting data at the same time as the rising and dropping of clocks, but the biggest difference is that DDR2 memory can be read 4bit. Twice the 2BIT prefetching of standard DDR memory, which means that DDR2 has twice times the ability to read the prefetch system command data for DDR, so DDR2 simply gets two times the complete data transfer capability of the DDR.
The biggest breakthrough in DDR2 memory technology is not so much the so-called two-fold DDR transmission capability, but in the case of lower calorific value and lower power consumption, instead, it gains faster frequency elevation, breaking the 400MHZ limit of standard DDR.
The difference between DDR2 and DDR
The main improvement in DDR2, compared to DDR, is that it provides twice times the bandwidth equivalent to the DDR memory when the memory module is the same speed. This is done primarily by using two DRAM cores efficiently on each device. In contrast, the DDR memory on each device can only use a dram core. Technically, there is still only one DRAM core on DDR2 memory, but it can be accessed in parallel, processing 4 data per access instead of two data.
Schematic diagram of the difference between DDR2 and DDR
Combined with the dual-speed data buffer, the DDR2 memory handles up to 4bit of data per clock cycle, one times higher than the 2bit data that traditional DDR memory can handle. Another improvement in DDR2 memory is that it replaces the traditional TSOP approach with FBGA encapsulation.
However, while DDR2 memory uses the same DRAM core speed as the DDR, we still have to use the new motherboard to match DDR2 memory, because DDR2 physical specifications and DDR are incompatible. The first is that the interface is different, the number of pins in the DDR2 is 240 stitches, and the DDR memory is 184 stitches, and secondly, the VDIMM voltage of DDR2 memory is 1.8V, and the 2.5V of DDR memory.
Definition of DDR2:
DDR2 (Double Data Rate 2) SDRAM is a new generation of memory technology standards developed by the JEDEC (Joint Commission on Electronic Equipment Engineering), which is the biggest difference from the previous DDR memory technology standards, although it was adopted in the rise of the clock/ The basic way of data transfer at the same time, but DDR2 memory has twice times the previous generation of DDR memory pre-read capability (that is, 4bit data read prefetching). In other words, DDR2 memory each clock can read/write data at 4 times times the speed of the external bus, and can run at 4 times times the speed of the internal control bus.
In addition, because the DDR2 standard stipulates that all DDR2 memory is in FBGA package form, and different from the current widely used TSOP/TSOP-II package form, the FBGA package can provide better electrical performance and heat dissipation property, It provides a solid foundation for the stable work of DDR2 memory and the development of future frequency. Recalling the development process of DDR, from the first generation of DDR200 to the PC through DDR266, DDR333 to today's dual-channel DDR400 technology, the development of the first generation of DDR has reached the limit of technology, it is difficult to improve the memory speed by conventional methods With the development of Intel's latest processor technology, the front-end bus requirements for memory bandwidth are increasing, with higher and more stable operating frequency of DDR2 memory will be the trend.
The difference between DDR2 and DDR:
Before you learn a lot about DDR2 memory, let's look at a set of data for DDR and DDR2 technology comparisons.
1, Delay problem:
As can be seen from the table above, at the same core frequency, the actual operating frequency of DDR2 is twice times the DDR. This benefits from the ability of the DDR2 memory to have twice times the standard DDR memory 4BIT prefetching. In other words, although DDR2 and DDR have adopted the basic method of data transmission at the same time as the rise and fall of the clock, the DDR2 has a capacity of twice times the pre-read system command data of DDR. That is, at the same 100MHz operating frequency, the actual frequency of the DDR is 200MHz, while the DDR2 can be 400MHz.
Then there is another problem: in the same working frequency of DDR and DDR2 memory, the latter's memory latency is slower than the former. For example, DDR 200 and ddr2-400 have the same latency, while the latter has a higher bandwidth. In fact, ddr2-400 and DDR 400 have the same bandwidth, they are all 3.2gb/s, but DDR400 's core operating frequency is 200MHz, and ddr2-400 's core operating frequency is 100MHz, which means ddr2-400 has a higher latency than DDR400.
2, packaging and calorific value:
The biggest breakthrough in DDR2 memory technology is not so much what users think of as two-fold DDR, but in the case of lower calorific and lower power consumption, DDR2 can get faster frequency increases and break the 400MHZ limit of standard DDR.
DDR memory is usually in the form of TSOP chip package, this package can work well in 200MHz, when the frequency is higher, its too long pin will produce high impedance and parasitic capacitance, which will affect its stability and the difficulty of frequency improvement. This is why the core frequency of the DDR is hard to break through 275MHZ. and DDR2 memory is in FBGA package form. Different from the current widely used TSOP package, FBGA package provides better electrical performance and heat dissipation, providing a good guarantee for the steady work of DDR2 memory and the development of future frequencies.
DDR2 memory uses 1.8V voltage, compared to the DDR standard 2.5V, reducing a lot, thus providing a significantly smaller power consumption and smaller calorific value, this change is significant.
New technologies used by DDR2:
In addition to the differences mentioned above, DDR2 has introduced three new technologies, OCD, ODT, and post CAS.
OCD (off-chip Driver): The so-called off-line drive adjustment, DDR II through the OCD can improve signal integrity. DDR II makes both voltages equal by adjusting the resistance value of the Pull (pull-up)/Pull-down (Pull-down). Use OCD to improve signal integrity by reducing the tilt of the Dq-dqs, and to improve signal quality by controlling voltage.
Odt:odt is a built-in core of the end resistor. We know that the motherboard using DDR SDRAM requires a large number of end resistors to prevent data line terminals from reflecting signals. It greatly increases the cost of manufacturing the motherboard. In fact, different memory modules do not have the same requirements for the end circuit, the size of the final resistor determines the signal ratio and reflectivity of the data line, while the end resistor is low in the data line signal, but the signal-to-noise ratio is also lower; the end resistor is high, the signal-to-noise ratio of the data line is higher, but the signal reflection also increases. Therefore, the terminal resistor on the motherboard does not match the memory module very well, and it will affect the signal quality to some extent. DDR2 can be built according to the characteristics of the appropriate termination resistance, so as to ensure the best signal waveform. Using DDR2 can not only reduce the cost of the motherboard, but also get the best signal quality, this is the DDR can not match.
Post CAS: It is set up to increase the utilization efficiency of DDR II memory. In post CAS operations, CAS signals (read/write/command) can be plugged into a clock cycle following the RAS signal, and CAS commands can remain in effect after the additional delay (additive latency). The original TRCD (RAS to CAS and latency) was replaced by Al (Additive latency), and Al can be set in 0,1,2,3,4. Since CAS signals are placed on a clock cycle after the RAS signal, the Act and CAS signals will never collide.
In general, DDR2 adopted many new technologies, improved the many deficiencies of the DDR, although it currently has a high cost, slow delay can be a lot of deficiencies, but I believe that with the continuous improvement and improvement of technology, these problems will eventually be resolved