Dude, even if the line was straight, with no curving whatsoever, and there was absolutely no processing delays in the network, and assuming the whole line is fiber optic, and it's incredibly efficient so as to allow the maximum angles possible <180 degrees in order not to drop effective speed, and c = 300000km/s, you would get 10ms 1-way. Ping is 2-way, thus 20ms alone for light, except if you were using neutrinos to transfer data and those weird experiment results actually turned out to be correct...