Introduction to Double Date Rate (DDR) Memory

In many systems, the advantages of serial data transfer over parallel data transfer are enough to justify the addition of circuitry that serializes and desterilizes parallel data in order to transmit it as serial data. However, one application area where parallel data transport is still extensively used is computer memory. Parallel interfaces are faster because they can read and write several digital signals at once, and designers are actively looking for methods to make them faster.

Double pumping, an antiquated but still effective method for achieving a significant boost in data-transfer rate, is what transforms a memory system into a double data rate (DDR) memory system. A memory system that utilizes “single pump” use the tried-and-true approach of synchronously transporting data from one IC to another. In essence, logic signals are sampled by the receiver on either the rising or declining edge of the clock, changing once every clock cycle. The word transfer rate in this system is restricted by the frequency of the clock.

For instance, you can (ideally) transport 133 million words per second if your clock runs at 133 MHz. It ought to be noted that because different computing systems employ different word lengths—8 bits, 16 bits, etc.—the bit transfer rate is depending on the system architecture. With a 133 MHz clock, a double-pumped system could accomplish 266 MT/s (millions of transfers per second).

In this post, we’ll examine the distinctions between single and double data rate interfaces as well as the benefits and applications of DDR.