It is often inefficient to buy into each generation of a product. If we already have one, we tend not to purchase the next until there is significant improvement. We wait for two or three levels of improvement before it seems cost-efficient.
Most people tend to wait 4-12 years between car purchases since there is no need to purchase a 2012 if you already own a 2011 or 2010.
Same thing with video cards and processors, I need at least two stages of improvement before rationalizing the need to upgrade.
We already have the
Fermilab Accelerator, with plans to help design the next generation accelerator/collider.
http://www.fnal.gov/faw/future/fermilabs-future.shtml
The next-generation particle collider
In collaboration with industry and DOE national laboratories, Fermilab is developing superconducting acceleration technologies. These SRF technologies have future proton and electron beam applications, including Project X and next-generation energy-frontier particle colliders. Scientists are developing proposals for several future colliders, including the International Linear Collider and a muon collider (see animation). Discoveries at CERN's LHC will soon determine which direction will best advance energy-frontier research. Beyond 2015, U.S. scientists will continue to make crucial contributions to the CMS experiment at the LHC thanks to Fermilab's Grid Computing Center and LHC Remote Operations Center.
Also, by geographically separating the research centers, the scientific community allows greater exposure to research opportunities for a larger number of people, while simultaneously distributing operation, control, and funding.