Living in Perth....
Ya wouldn't know it.
If you want to bring data from Europe or Asia to Australia, routing it to the city of Perth on the nation's west coast is a good idea because the bits spend less time on a submarine cable. For traffic from the USA, routes to Sydney on the east coast are similarly sensible. Once you land that data you then have the chore of …
True. but traffic within Oz is often OK. The problem is offshore data. Even the onshore sites tend to load their website pages with advert and framework stuff that has to come from halfway around the planet. So does that mean that our experience in Perth will be further degraded to support Melbourne and Sydney?
I was going to moan about how slow 40ms is for that distance, then I checked the distance - I forget how big Australia actually is.
Light travels go at the speed of light (funny that), but bounces all round the fibre instead of going straight, roughly add 50% to the distance to account for that, giving a total travel time of approximately 20ms.
The distance is probably longer than I've calculated, fibre doesn't travel as the crow flies, but its in the ballpark. For comparison, London to NY is routinely about 60ms and is about 40% further (5500 km vs 3900 km).
ish.
If I have my maths right...
Light through fibre is usually around 2/3 the speed of light in a vacuum - around 204,000 km/s versus 299,792 km/s). That gives a best case of 54ms for a round-trip between London and NYC and 38ms for Sydney to Perth.
I'd expect NYC-London to be slightly higher (although it is now possible to send traffic without repeaters, I would still expect them to be in use on current circuits and the distance is the direct route versus the actual cable distance) while the Sydney-Perth route is probably accurate based on the distance between the two cities via the A1 highway.
I guess they are deploying 100G technology which means each wavelength on this fiber path provides 100G and usually such a cable would feature dozens of fibers each being lit up with at least 48 wavelenghts. So the total capacity is x fibers * n wavelenghts * 100G and not only 100G.