Aaah memory. It has been some time since we last had a memory review, and for good reason. Memory got pretty boring. Ten years ago this was not the case. DDR was just fresh on the scene and we were starting to see memory speeds and bandwidths get to a place where it would have a significant effect on performance. Latencies were of utmost importance, and the fastest 184.108.40.206 DIMMs running at DDR 400 speeds were often quite expensive. Then things sort of mellowed out. DDR-2 did not exactly bring faster performance over DDR initially, and it was not until DDR-2 800 and 1066 speeds that we actually saw a significant boost over previous gen DDR 1. DDR-3 brought even more yawns. With the jump to integrated memory controllers from both AMD and Intel, DDR-3 speeds were nearly meaningless.
The primary reason for this rather vanilla time in the memory market was that of individual bandwidth needs for CPU cores. Most research into this issue points to an individual CPU core needing only 3 to 4 GB/sec of bandwidth to support its data needs. AMD and Intel have gone to great lengths to increase the efficiency of not only their memory controllers and prefetchers, but also the internal caches so fewer main memory accesses are needed. So essentially a quad core processor would really only need upwards of 12 to 13 GB/sec of bandwidth in real world scenarios. DDR-3 1333 memory modules in a dual channel configuration would be able to support that kind of bandwidth quite easily. So what exactly was the point of having faster memory? Also, CPUs using DDR-3 memory are not as sensitive to latencies as we have seen in previous generations of parts.