Remarkably, it shows a cable owned/installed by Zayo that goes right through the edge of my tiny village in rural New Mexico. The company, AFAICT, is not consumer-facing - too bad, because if it was, it would be a local gamechanger.
Look at the UK for example, apparently "iomart DC4 Leicester" is just off Berkeley Square, London.
Not only is that technically incorrect (Leicester != London), but its hilarious to propose that anyone would build a datacentre in that part of London (its hedge-fund central where rents are three-digits per square foot .... plus there isn't the space anyway to build a datacentre of any sort of meaningful scale).
And the map is just full of such examples.
So yeah, I've grave concerns for the accuracy of that resource.
Interesting. I was looking at the seemingly odd choice of a loop out of Texas to Mississippi through the Gulf. Then I realized the roughnecks on the drilling platforms need their netlfix bing fixes too!
This is literally the article that resulted in my mid-1990s teenage self being motivated to work in ISP core network engineering, which is what I've been doing now for twenty years...
I wish Stephenson would write more things like this in addition to his novels.
I wonder if there is a version of this that includes domestic cables? Navigational charts generally show them (as well as various "DO NOT ANCHOR" signage denoting them). It would be cool to see what the more "minor" or "non-international" underwater infrastructure looks like.
Is there a version of this that includes the bandwidth provided by each cable? What is the total cable bandwidth between the East Coast North America and Europe?
I'm guessing the maximum possible throughput of the cables themselves is significantly larger than the actual throughput or even the maximum throughput of the whole system taking into account the equipment on each it.
If you have suggestions, please simply click the wrong / error link/datacenter and submit correction. This is a community built, and managed platform, it will only get better if you help :) thank you
In the US there are quite a few landings just outside of NYC and LA. I'd guess London is too far inland, and it's cheaper to run fiber over land.
There is probably also an incentive to keep the cables away from major shipping lanes where they'd disrupt the ability for ships dropping anchor. Notice around Seattle some cables snake up and through the sound, but it's also extremely deep water. That landing is also far north of the city proper, so any ships waiting near port are a ways from the landing site.
Just in 10 years South East Asia is full of submarine cables now. These days you can easily get 100mbps fiber optics subscriptions in some village in an island there, which is a huge upgrade considering you can't even get a landline 20 years ago in those remote areas.
I was on holiday in Bude, Cornwall, UK this summer. I’d have taken the family on a day out touring cable landing sites if I’d have known there were 5 nearby !
What do they actually look like though ? I’m guessing it’s not a shed at the top of a beach, but something more robust ?
A lot of the time these landing sites / buildings are a bit further inland. At the beach there is usually only a buried armoured cable leading to a concrete slab or something where it connects to underground ductwork.
If you're interested in that and have an hour of time there is a youtube video[1] of a local tracking these down around the Cornwall area.
The cable landing I’ve seen outside Skagway, AK is just a shed on the rocks. And about a hundred warning signs of the federal crimes you’re committing if you get close.
Every time I see this map I think to how much Europe is dependent on US Internet services (email, cloud and many more) passing through those cables and how easy it would be for a hostile actor to block the European economy using the Nord Stream technique.
I work up at the north of Hudson's Bay in Québec, where we're all eagerly awaiting the arrival of the fibre connection. For those curious, all cables in grey seem to incomplete projects. For instance, the Kattittuq Nunavut cable only just got funding.
If you want to learn more about the first transatlantic cable check out How the World Was One by Arthur C. Clarke. Reading it now and really enjoying it: https://amzn.to/3EiACjU
Does anyone know how the cables get from Hillsboro (OR) to the ocean? It appears to be the most in-land landing point on the map. Along the right-of-way for US 26?
From the Oregon coast landing stations, one of the predecessor entities acquired by level3 (now lumen) built a ring of fiber to Hillsboro, another was built by wave (now known as rcn/grande/astound) much later with as much diversity as they could from the existing route.
Taiwan biggest trading partner is, by far, mainland China. And Hong-Kong is not far down the list [1]. Media might give you the impression that China and Taiwan are practically at war, and that might even be true in so far as China is contemplating invading Taiwan.
But if you look at the history between the Taiwan and China [2], there has been a lot of economical cooperation, after the Tiananmen Square Massacre in 1989 (many foreign companies left China at the time, leaving a void filled by Taiwan), and especially between 2008 and 2016. So it's not surprising at all that there are cables in the strait.
The length of it is exaggerated on the map due to map projection (obligatory comment). In reality it's about as long as the one connecting Oregon to Taiwan (which is long, but not record-setting.)
Probably a pretty important project to do that one for Russia anyway, it connects all parts of the country.
That made me think that laying underwater cable must be very economical. There's lots of loops that look completely unnecessary, like the Gulf of Mexico. But maybe it was cheaper than installing cables on land? And the terrain of Northern Russia is probably a lot harder to deal with than Texas.
Cheaper but typically a longer route which translates to higher latency.
As a standard internet user the difference between 20ms and 24ms might not matter to you but when selling enterprise services, especially around financial institutions (but not exclusively) that 4ms matters. Often it matters enough that the business case for terrestrial cable starts to make sense.
> But maybe it was cheaper than installing cables on land?
Way, way cheaper. Trenching or tunneling on the one hand, mostly just steaming forward whilst paying out cable on the other. The cable itself is more expensive for submarine use though, so there would be a point where cost crosses over.
https://www.infrapedia.com/