Dual Channel Vs. Quad Channel

[Muchobanana]

Currently i am using an Asus Z87-A mobo which has a dual channel architecture. I am using some vengeance ram (1x8gb) 1600MHz which says it is quad channel ram. This was my first build and so i was no expert when i built it. My question is will i notice a difference if i were to replace the 8gb stick with an 8gb dual channel kit, maybe some 2133. Alternatively, could I add an identical 8gb stick to the one i have now and would it run in dual channel even though the specs say its quad channel ram. Which option should i choose, or neither and i can keep the current stick of ram? Thanks for the help in advance.

 

[boju]

Memory can be fussy with each other (miss match) and may cause system problems. When people buy 2 or more stick of ram, besides being convenient, they opt for kits of 2 or 4 so the memory that is packaged are compatible with each other. So Dual or Quad channel labeling is just that, a marketing label.

Memory miss matching is quite rare these days, long as all memory has the same specs.

Your motherboard supports dual channel so buying another 8GB 1600MHz would be the way to go if its the cheaper option. You wont notice any different between 2x 1600MHz or 2x 2133MHz. But your system will perform better with dual channel memory.

If you buy another memory singly, make sure it has the same specs, voltage, timings etc.

 

[Tradesman1]

First off a single 1x8GB stick isn't quad channel - a single stick is single channel (and DRAM isn't 'made' as Single, Dual, Tri or Quad channel), DRAM is DRAM, each stick is a single 64bit device. When you have 2 or 4 sticks running in dual channel, the memory controller sees all of the DRAM as a single 128bit device which can result in a 10-15% performance increase....if we add your thoughts of kicking up to 2133 sticks then yes you could see a nice improvement - while gaming might show a 1-3 FPS increase, your real gains would be when multi-tasking, doing imaging, video, VMs, CAD as well as other memory centric apps and/or using large data sets