Unfortunately, latency is the hardest and slowest of all network properties to fix. Part of the problem stems from the fact that, as mentioned above, very few services and applications require ultra-low latency delivery. This limits the business case for any network operator or latency-centric content delivery network CDN—a business case that is already challenged and conflicts with basic laws of physics.
It takes 40-45 milliseconds of light to travel from New York austria mobile database City to Tokyo or Mumbai over a distance of 11,000-12,500 km. This meets all low latency thresholds. However, while most of the Internet backbone is fiber, fiber cable is about 30% slower than the speed of light because it is rarely in a vacuum +losses are typically 3.5 dBkm. Copper and coaxial cables experience more severe latency drops over long distances and have more limited bandwidth, which means there is a greater risk of congestion and delayed deliveries. These cables still make up the majority of the cables inside residential and commercial buildings and in neighborhoods.
Furthermore, none of these cables are laid in large numbers. And what we typically think of as the “Internet backbone” is actually a loose consortium of private networks. As a result, the network distance between a pair of servers or a server and a client can be much greater than their geographic distance. Furthermore, network congestion can cause traffic to be routed less directly to ensure reliable and continuous delivery rather than to minimize latency. This is why the average latency from New York City to Tokyo is more than 4 times the time it takes for light to travel between the two cities, and the average latency from New York City to Mumbai is 4-6 times.