Broadband Speed Test

We all know that there was an awful earthquake on Boxing Day last year that packed a rather big wallop against our internet use. But do you know that out of the 13 submarine cables over there that provides Singapore with a direct linkage with America's servers, only 2 have been repaired to date?

Speed test results with a local server.

Speed test results with a server based in Yokohama, Japan.



Speed test results with a server based in Los Angeles, the U.S.

Why is the U.S. such a crucial destination you may ask? While it appears that the growth of the Internet was fuelled by generated web content which is diversified in terms of area of origin, the truth and the more important point is that much of these contents are hosted by powerful blade servers that reside primarily in the U.S. Another important point to note- despite its sprawling size and outreach, the Web actually relies on 7 core DNS servers that assist in translating domain addresses into PC-readable IP addresses for communications between clients and servers to work properly. Again these 7 DNS servers reside in the U.S.

It's interesting to note that the program which I employed for the purpose of finding out my downloading/uploading speeds, deliberately routed me to the test servers in the U.S. via the Middle East instead of the Pacific Ocean. Which proves my point that the 13 submarine cables are still out of order. At the same time, I would also like to highlight the fact that the touted broadband speeds by your friendly ISPs in reality apply only for local WANs covered by their infrastructural assets. In other words, once we are out of the network of cables under Starhub/Singnet, our speeds are at the mercy of companies who own the cables that link our local networks with the networks hosting the servers (which we want to connect to). A classic case of a bottleneck is what we are seeing now.

It's surprising that actual hardware support for increased data bandwidth hasn't grown much despite the exponential growth of demand for greater and faster data access. In fact, investments in submarine cables largely remained stagnant by the mid 1990s. How the ISPs will handle increased demand for higher bandwidth will remain to be seen, for the earthquake incident has exposed one key weakness of the Web- her reliance on too few critical access nodes.

P.S. I wonder what's the hullabaloo over the implementation of a gigantic wireless network in Tai Chung... Is that investment a sure case of targeting actual demand? Heh. Sadly I believe they have missed the point totally, period...

P.P.S. Sorry but I just can't resist taking potshots... :P