Subsea Cables & Internet Infrastructure

This article shows that if Starlink deploys a massive network of its 1.5 terabit satellites, it yields nothing better than 83 megabits throughput per square kilometer. In contrast, ITU 5G standards require 5 megabits minimum per m2. Moreover, water covers 71% of the Earth, which guarantees massive satellite excess capacity, and consequently, long term price wars as the Chinese, Europeans, and Amazon enter the market. The earth's surface exceeds 500 million square kilometers. Moreover, LEO satellites must be evenly distributed in orbit to ensure global coverage. The result is that a 1 terabit down Starlink satellite does not carry as much punch per kilometer as enthusiasts and groupies believe. A 42 petabit satellite constellation divided by 500 million square kilometers equals 10 bits per meter.  Doesn't look promising to me.  https://www.aethaconsulting.com/leo-satellites-and-terrestrial-networks-competitors-or-complements/

Hi Everyone,  I am available to help you acquire network capacity based on your requirements. In general I don't charge because the carriers will pay me for any subsequent sales. So if you work with me and you buy, then you will contract directly with the network operator. This ensures you get the best possible price. My value-added is my knowledge of the operators, market pricing, delivery performance, latency, uptime, commercial flexibility, and the like. I am also good at negotiating deal terms.  In a few cases I take a retainer because the client is looking for long term consulting.  You can reach me at roderick.beck@networksourcing.net. I live part of the year in Europe and also in Africa. 

This 24 fibre pair cable connects Mombasa, Kenya to Oman. It is a joint project of META and Safaricom (Kenya's largest mobile provider). Safaricom is the Mombasa landing partner, but META is the cable's sole owner. ASN is the construction partner. Capacity throughput should be a half petabit per second. Safaricom's participation reflects a long term trend for large mobile operators to enter the subsea cable business. Other examples including Vodafone, Bharti, Ooredoo, Axiata, and their subsidiaries. Subsea cable ownership allows mobile operators to expand operations overseas and also reduce the operating expenses for their international voice traffic. Oman is rapidly becoming a subsea cable hub as it offers an appealing alternative to the United Arab Emirates. The UAE's high cross connect CLS charges constitutes backward protectionism designed to protect its two telecommunication incumbents. Secondly, the UAE is currently simply too close to Iran to be a good place to l...

AI data centers has two challenges: big capex requirements and high operating expenses due to the skyrocketing cost of electricity. Another challenge is low customer switching costs. A customer like OpenAI or Anthropic rarely puts equipment into an AI data center. It simply downloads its software into the data center's bare metal servers, estimates its model, and then uploads the estimated model back to its own servers. Oracle has only one major customer and so it faces huge customer risk. Contracts tend to be soft and often include no penalty cancellation clauses. This business model only works if the vendor has access to cheap electricity like hydropower in the Nordics and very low customers. Solar and wind sell rates do not generally make the grade because fully loaded they are in fact quite high. In the case of Oracle, it went from a typical debt free software company to a 2.4 debt to revenue ratio. This includes bonds, long term lease liabilities, and short term obligations. 

Not only do I offer Equiano 100Gs below $20K with you contracting directly with the network operator, but also available is Layer 2 50G protected MPLS circuits on Equiano for $13,500 MRC on two year terms. This allows you to migrate from granular Layer 2 to Layer 1 as your business scales. African ISPs need to seize the day before capacity tightens and the window of opportunity closes due to the lack of any new cables for 3 to 5 years.

Singapore currently has 26 operating fiber optic cables. The government is aiming to double that figure over the next ten years. It is also wants to increase residential broadband to 10G, which seems frankly far fetched and perhaps overkill. The biggest drawbacks to landing in Singapore is the limited number of cable landing stations and the congested sea harbor. I would tell the government that it needs more cable landing stations and that they must be physically diverse to the existing facilities. More Malaysian/Singapore terrestrial crossings for Trans-Continental cables, but that conflicts with the sovereignty angle dominating policy making today. https://www.edb.gov.sg/en/news-and-insights/singapore-to-double-capacity-for-subsea-cable-landings-to-enhance-digital-connectivity

1. Fully diverse and fully buried dark fibre pairs or individual strands available for lease or IRU between IColo1, IColo2, Telehouse, Nyali Exchange, Seacom CLS, and 2Africa Shanzu CLS. Unlike other providers, the fiber connects to the data centers via buried conduits. Not power poles. End to end burial. 2. Sixty Euros per kilometer on three year lease deals. 3. New ultra-loss fiber manufactured in 2024. 4. Repair SLA based on 4 hours mean time to repair.

 At the 2026 Submarine Networks World in Singapore Ciena pitched the MOFN business model, the managed optical fiber network. This is a network dedicated to a single customer like a hyperscaler, large enterprise, government, and the like. It involves dedicated fibre pairs and equipment. Very often the customer buys the equipment and ships it to the network operator. It is quite common in Africa where network operators provide such services to the American Tech Giants. But carriers have often found that hyperscaler forecasts of capacity needs are often wildly off. The result is poor MOFN economics and low rates of return.

Ciena seems correct when it asserts massive growth from here to 2035. Heck, that is a 35 year Internet trend. But claiming that traffic will be 85% AI sounds like a very bad forecast. Right now the world is spending $2 trillion each year on AI infrastructure versus $100 billion in AI service revenues. A pretty big gap. We have experienced cycles of excess technology enthusiasm over the last 150 years and bullish forecasts have almost always proved wrong. 

The second Kenyan data center (IXAfrica Nairobi) I've visited with a large Starlink presence. IXAfrica is on Nairobi's outskirts on Mombasa Road and is pretty damn difficult to find as I and the taxi driver learned to our dismay. 😃 Starlink normally requires 20 degrees of line of sight in all directions for its earth stations. A condition clearly violated here. I suspect Starlink is using free space lasers to transmit traffic to satellites high enough in the sky to reach this limited-line-of-sight facility. From my point of view all this talk about Starlink building a parallel Internet infrastructure qualifies as nonsense. It is just another niche ISP using a different and bandwidth-limited Last Mile infrastructure to reach end users. No revolution. Just hype.

Subco announced recently four new 24 fibre pair cables connecting Australia to Japan, Singapore, Guam, and the Continental US. This is obviously a multi-billion dollar undertaking. Now Subco owns the Oman-Australia and SMAP cables. It is also a consortium member on Indigo West and Indigo Central. So I believe Subco has significant retained earnings. Nonetheless, these new cables together probably come close to the $2 billion construction dollar mark. Moreover, hyperscalers don't buy capacity prior to successful execution. Indications of interest, yes, but not contractual commitments. LOIs are soft commitments. So this is a daring undertaking by one of the most successful private operators on subsea cables. For those new to the industry, a private operator generally is a wholesale player that owns an entire cable as opposed to a consortium member or a hyperscaler. However, the historical track record for private operators has not been great. So far Subco has an exception. By 2001, p...

Looks very similar in design to our present fibre optic subsea cable networks. For example, the Red Sea is a telegraph cable highway and Singapore is a telegraph cable hub. However, the Persian Gulf is not important because hydrocarbons won't be discovered there for another 30 years. Note the cable connecting South Africa to Australia, which is reminiscent of Google's Umoja cable. Note the cables connecting Argentina and Brazil to Senegal and South Africa, but Nigeria, a major hub today, has no connectivity.

I came across several instances of dubious subsea cable business models at the conference. 1. For example, one subsea cable startup thinks it can skip equity investments and actually finance its project via debt in the form of bonds. This idea makes absolutely zero sense to anyone with a finance background or experience in the field. The problem with debt is that you must make payments even through your revenues will be zero for several years while the project is executed. Furthermore, it is total crap to claim you will sell capacity IRUs prior to RFS. The simple fact is that sales prior to substantially finishing the project never happen. In the six months prior to RFS sales can happen because carriers can see the project is almost ready. Indeed, that is when the hyperscalers sell capacity. But a debt-financed project needs big IRU deals from the day that bonds are issued. Financial history is quite clear on this point. Telecom infrastructure projects in their early stages are almost ...

This is the Peace cable's Marseille backhaul. It illustrates standard landing architecture as well as good network design. First of all, the cable comes ashore outside Marseille Port's own cable landing facilities for the sake of physical diversity and resilience. Always put some distance between a cable and others if possible. This increases the cable's value and demand for its capacity because diversity means adding it to the network portfolio improves its resilience. It is analogous to the role of diversification in a financail portfolio. Like all other subsea cables, Peace terminates in a beach manhole where it was spliced to a terrestrial fibre cable that also contains a power conductor just like the subsea cable. Then the terrestrial fiber goes to an Orange 'CLS', really just a power feed hut. From there the cable becomes part of a fully diverse fibre ring that includes dual entrances into MRS2 where most of the cable's SLTEs are kept. Most likely this fib...

Route Latency SG/Marseille 136 ms Mombasa/Marseille 104 ms Karachi/Marseille 92 ms Karachi/Signapore 82 ms Kenya/Singapore 93 ms.

Everyone at the conference wants to be Keppel. The carrier bought five fibre pairs on the 12 fibre pair Bifrost subsea network at construction cost and then sold them individually for a total of $1.2 billion dollars. This is a classic example of bandwidth arbitrage that has investors outside the industry salivating and has many Pacific wholesale carriers hoping to do the same thing in the future. Why do the heavy lifting of providing lit services when you can buy fibre pairs at construction cost and then flip them? What made Keppel so successful is that Bifrost was the first cable to directly link Asia's premier telecom hub, Singapore, to the US. And the fact that the hyperscalers kept the rest of the capacity for themselves. So Keppel ended up as the monopoly supplier of capacity to the entire wholesale market. So the moral is be the sole carrier partner on a sexy hyperscaler project and watch customers coming running to you with the enthusiasm of teenage girls at a Beatles concer...

Most African cables traverse the Congo Canyon as opposed to detouring around it. The canyon is not particular deep, but it is huge in terms of floor coverage. Depth varies from 1400 to 1600 meters. Le Trou Sans Fond canyon off Abidjan drops 2300 meters where the canyon meets the continental shelf. It is perfect recipe for sediment and debris avalanches.  These canyons were created by strong river currents that carve the ocean bottom and also deposit sediment. Debris and sediment slides can easily crush the older African cables lying at the bottom. In contrast, both Equiano and 2Africa are routed around these danger spots. That increases latency, but reduces outages. Sometimes African ISPs are short sighted in this matter. They take WACS despite the exorbitant costs and high likelihood of extended down time because it is a bit faster.  An interesting question is how long will WACS be worth operating given the accumulated optical loss from one or two outages each year. The more ...

My understanding is that standard practice is to slice a damaged cable in two prior to retrieval. Hard to lift a 14,530 kilometer cable otherwise. 😃

 National security analysts please note, there is no sabotage category because there are no credible cases of sabotage. The IPCC study estimates 3,224 subsea fibre optic cable repairs over the last 18 years. It classifies 84% of outages as caused by fishing and anchors, 5% due to abrasion, 4% due to plant failure (subsea optical amplifiers or PFEs), and 7% due to geological disturbances such as turbidity currents induced by earthquakes or heavy rains or debris slides in subterranean caverns. Abrasion often leads to shunt faults where water reaches the power conductor.

Probably not. Right now carriers are charging $7K to $9K per month for a 100G wave linking IColo1 and IColo2. This is the pricing of a cozy oligopoly, not a competitive market. Moreover, the carriers have inflated their costs by using brand name hardware when comparable products like Smart Optics offer 50% to 60% discounts on both the price tag and operating expenses (there is good Layer 1 networking gear with no recurring costs in the form of port activation fees or licensing). I can refer you to a smart optics VAR. In contrast, you can lease today recently manufactured dark fibre pairs between IColo 1 and IColo 2 for $1,100 per month for the 18 kilometer primary path and $1,400 for the 23 kilometer backup path. So for $2,500 a month on a 3 year contract you can get a fully diverse fibre ring between these two important data centers. Moreover, it is unlikely you need long range optics. Each fibre pair can do 10 to 20 terabits.