CPU clock rates have experienced an exponential growth in the last few years. This growth has left the rest of the PC components behind. The resulting high-end systems are deprived of better performance because of the lag that other components create. The memory bus constitutes what most consider to be the greatest problem. Ramping up bandwidth only partially solves the problem that the bus causes. Latencies become the primary clock rate reduction problem. Reducing latencies by means of faster strobes is difficult and economically not viable. Alternative solutions encompass combined SRAM-DRAM that mask latencies by uncoupling data output from the DRAM array. A simple row cache architecture, using time multiplexed internal buses to load entire rows into centrally located 8-kb SRAM cells, can function as an output buffer. Thus, the DRAM array can be excelled ahead of time to avoid page closing latencies as well as refresh penalties.

In 1997, SD RAM appeared on the market. SD RAM was an alternative to DRAM and SRAM. It was able to run at a faster clock rate. As a result of a much simpler communication mode, that SD RAM employs, all commands, addresses, and data are timed to a single clock signal. SD RAM is a burst mode memory that bursts a series of data words at the 66-MHz clock rate following the initial 58-ns random access latency. The typical organization of the memory modules changed from the single in-line memory module (SIMM) to the dual in-line memory module (DIMM) that supported the entire 64-bit databus width with a single module.

In 1998, the SD RAM frequency had increased to a 100-MHz burst rate. The random access latency improved to 56 ns. Both processor clock speed and memory bandwidth had increased by 12 times, but the DRAM latency in clock cycles had become more than 5 times worse. Computer system performance started to become limited by latency and not by bandwidth.

In 1999, corporate competition between Intel and AMD increased and so did the CPU clock speed. Unfortunately, this rapid development in the processor industry, created a further widening of the cleft between CPU clock speed and the rest of the system components. This led to the increase in memory bus speed to 133 MHz. Yet, the chipset still suffered from latency problems.

In order to deal with this problem, DDR RAM (double data rate transfer) was designed. At first, DDR RAM was primarily used in graphics cards. Because you only need to use 32 MB to get 64 MB of performance, many companies were using this to cut down on their expenses. AMD is the first company to use DDR RAM in their motherboard.