The question arises as to whether ketanrd01 requires a square wave or a TTL compatible signal.
The G series TTL (linked to both by myself and Daniel) would clearly produce a relatively fast rising and falling edge but it is interesting that the
PO74G14 can do 1.125GHz into a 2pF load, but that the input capacitance of the gate itself is 4pF!
Of course it can do up to 100MHz into 50pF, so you're unlikely to hit that limit, but I wonder how often you'd see these running at the headline 1.125GHz?
On the very first point I raised, I have a 10MHz rubidium frequency standard with a TTL compatible output, and even that looks a lot more like a sine wave than a square wave.