What is the difference between SDRAM and DDR SDRAM? Which performance is better?

Synchronous dynamic random access memory (SDRAM): It is a widely used memory type in PC 100 and PC 133 specifications. At present, its bandwidth is 64 bits, the voltage is 3.3V, and the highest speed of the product can reach 5ns. It uses the same clock frequency as CPU for data exchange, and its working frequency is synchronized with the external frequency of CPU, so there is no delay and waiting time.

Double data rate SDRAM (DDR SDRAM): DDR for short, because data can be transmitted at the upper and lower edges of the clock trigger edge, the bandwidth can reach 2. 128GB/s even at the bus frequency of 133MHz. DDR does not support LVTTL with 3.3V voltage, but supports SSTL2 standard with 2.5V It can still use the existing SDRAM production system, and its manufacturing cost is slightly higher than that of SDRAM, but it is still much lower than that of Rambus. Because the equipment for manufacturing ordinary SDRAM can produce DDR memory with a little improvement, there is no patent problem, so it represents the memory development direction that can compete with Rambus in the future.

Direct Rambus DRAM (DRDRAM) is the development direction of future memory advocated by Intel. RISC (Reduced Instruction Set) is introduced into it, and the data volume of each clock cycle is simplified by high clock frequency. The working frequency is higher than SDRAM (not lower than 300MHz), but the data channel interface bandwidth is low, only 16 bits. When the working clock is 300MHz, Rambus uses the upper edge and the lower edge of the clock to transmit data respectively, so its data transmission rate can reach 300×16× 2/8 =1.2 GB/. It is different from the traditional DRAM in that the pin definition will change with the command, and the same set of pin lines can be defined as address lines or control lines. Its pin count is only one-third of that of ordinary DRAM.

Virtual Channel Memory (VCM) is the memory standard supported by most latest motherboard chipsets. VCM is a new type of "buffer DRAM" developed by NEC Company, and this technology will be used in large-capacity SDRAM. It integrates the so-called "channel buffer", which is configured and controlled by high-speed registers. VCM not only realizes high-speed data transmission, but also maintains high compatibility with traditional SDRAM, so VCM memory is usually called VCM SDRAM. The difference between VCM and SDRAM is that data can be submitted to VCM for processing whether it is processed by CPU or not, while ordinary SDRAM can only process data processed by CPU, which is why VCM processes data more than 20% faster than SDRAM.