With the Core i7 platform, memory speeds of up to 1600MHz are officially supported, but kits are sold with speeds of up to 2133MHz and beyond.
Are the caveats/drawbacks of using faster-than-rated memory the same as overclocking, or is there a specific technical reason why the faster memory is unsupported or could cause problems?
Edit: I am referring to running the memory at the advertised module speed (2133MHz, for example) on a platform that doesn't "officially" support it. I know that the module would downclock by default and run at the normal supported bus speed without any intervention.
In servers, I realize that keeping the bandwidth in check is a critical issue, especially with dozens of DIMMs and hundreds of GBs flowing around. Perhaps these are related?