Dual-Channel Architecture
Contents
As we briefly discussed in the previous page, the dual-channel architecture expands the number of data wires available in the memory data bus from 64 to 128. This doubles the available bandwidth. For example, if you use DDR3-1333 memories, the maximum theoretical transfer rate is doubled from 10,664 MB/s (10.6 GB/s) to 21,328 MB/s (21.3 GB/s).
Each memory module, however, is a 64-bit device. Therefore, in order for the dual-channel architecture to work, you will need to install two memory modules in parallel, making 128 bits available.
Many people have trouble visualizing this idea. Therefore, let’s draw some schematics. First, assume that we have a system that doesn’t support a dual channel feature (i.e., a single-channel system). In this case, the memory controller will transfer 64 bits at a time. When we say that the memory data bus is 64 bits wide, this means that there are 64 wires (yes, physical wires on the motherboard) connecting the memory controller and the memory sockets. These wires are labeled D0 through D63. The memory data bus is shared amongst all the memory sockets. The address and control busses will activate the proper memory socket, depending on the address from where data must be stored or read. We illustrate this in Figure 3.
Figure 3: How single channel works
On systems supporting dual-channel technology, the memory data bus is expanded to 128 bits. This means that there are 128 wires connecting the memory controller to the memory sockets. These wires are labeled D0 through D127. Since each memory module can only accept 64 bits at once, two memory modules are used to fill the 128-bit data bus. See Figure 4. Because the two modules are accessed at the same time, they must be identical (same capacity, same timings, and same clock rate).
Figure 4: How dual channel works
Now that you know what dual channel means, the most obvious question is: “How can I enable this architecture so I can increase my computer’s performance?” Let’s find out.
