This article is part-two of the previous article – Mumbai-Chennai-Singapore, the Superhot Submarine Superhighway. It completes the narrative on new submarine cables to be unveiled in the course of 2020 that are set to land in India. Designating it part-two made sense as Mumbai-Chennai-Singapore initiatives is few steps ahead of Mumbai-Marseilles with respect to new submarine cables.

Having said that, the two subsea routes have deep synergy with respect to Marseilles-Mumbai-Singapore Eurasian express route. With Mumbai firmly positioned as the hub for new submarine cables, many never before developments are set to unfold, and challenge the status quo. Complimentary alliance for cable system interconnect and fiber pair swap will be on the table going forward, though some will explore new cable build on either of the routes as Phase-2. Meanwhile, new cable system build vs fiber pair acquisition will continue to be decisive determinant for OTT, Telco Carrier and IIP community and influence their subsea network acquisition roadmap.

Characterizing the subsea route

New Mumbai-Marseilles cable systems in making are heavy lift initiatives with long haul subsea route with multiple country cable landing in Middle East, North Africa and Western Europe. Western Europe has France and Italy as two submarine cable landing hubs positioned as gateway to Europe. While Marseilles has more prominence, potential and talked about frequently, Italy lands higher number of Mumbai-Europe cable systems in three cable landing points in Sicily. Tollgate market play of incumbent monopolies in North Africa adds exasperating dynamics and the much talked about exorbitant commercials. Easing geopolitical situation in Middle East opens new terrestrial-cross-border connectivity pathways leading to new cable route design and diversity options. Some of these routes will create new low latency benchmarks. Mumbai is poised to witness multiple cable landing and that makes selection of landing party for open cable landing and dark fiber backhaul a critical decision and important success factor. In all there are many moving parts to be synchronised for design and deployment of new cable systems.

India with its Asiatic affiliation has been part of APAC business unit of customers and service providers across all market verticals. In some cases, and particularly with respect to subsea network build and acquisition, Mumbai, not including rest of India, is considered part of EMEA region. This leads to unique positioning for Mumbai as logical midpoint for <EMEA>-<Mumbai>-<APAC> for submarine cables.

Mumbai-Marseilles is truly long-haul transcontinental subsea route and hence a challenging proposition for singular-ownership submarine cables. With respect to existing cables and also for some of the new cables, it has four segments – Mumbai-Djibouti (3,300 kms), Djibouti-Egypt (3,000 kms), Egypt Crossing (400 kms) and Egypt-Marseilles (3,300 kms). The subsea route of 10,000 kms makes new subsea cables nearly $450M ventures, nearly 2X of Mumbai-Chennai-Singapore cable systems. Hence to raise funding is that much more challenging and that possibly is the reason new submarine cable initiatives have taken longer to take off.

State of Consortium Cables

Owing to heavy investment needed and multiple countries enroute, Mumbai-Marseilles subsea route has been one of the most prominent consortium cable routes globally with SMW3SMW4IMEWEEIG and AAE-1 for the last 15 years and SMW6 going forward. BBG also joined the gang interconnecting with EIG in Barka, Oman, facilitated by Omantel that is consortium member in both the cable systems. The "deal" to make cross-connect charges fit-for-business must have been interesting. Understandably such interconnect couldn’t have happened in Mumbai. It would be unfair not to refer to SMW5, though it remains unexplained why did not land in India. These cable systems have played a great role in the 10G era. IMEWE and AAE-1 graduated to 100G and 200G era well on time upgrades and hopefully EIG also followed the suit, to address surging demand. They will remain relevant for longer, but with limitations of shared capacity pool and restricted to capacity-play.  

The consortium cables, however, are not architected to address terabit scale fiber pair ownership economics, where owners of individual fiber pairs and even fractional fiber pairs have the independence to design how the fiber pairs are lit and how it is extended inland for termination and capacity breakout. This change in dynamics of acquisition, ownership and operationalization of subsea connectivity is the driving force for new high fiber count cable systems. With IMEWE-2, EIG-2, BBG-2 or AAE-2 not in the horizon yet, the consortium cable systems are likely to become antiquated infrastructure with precipitous decline as the new cable systems reach RFS in 2022-23.

State of Privately-Owned Cables

The privately-owned cables and India West Coast counterpart of TIC and i2i are TGN-EA from Tata Communications and FALCON+HAWK and FLAG EA from GCX. Trailing behind is Airtel with subsea asset ownership in GBI/MENA and partnership with Telecom Egypt.

The 4FP TGN-EA has intriguing architecture. While Tata Communications built the Mumbai-Djibouti-Egypt segment with landing in Jeddah, for 4000 kms Egypt to Marseilles segment, Tata Communications along with SEACOM acquired 1FP each on TE North from Telecom Egypt. The branch, SEACOM, lands in Kenya, Tanzania, Mozambique and South Africa. SEACOM owns the East African submarine cable along with 2FP in TGN-EA between Egypt and India. Hence TGN-EA with respect to Mumbai-Marseilles remained a 1FP cable system.

FALCON+HAWK isn’t any better architecturally with HAWK initially built to connect Marseilles and Cyprus and the branch to Alexandria, Egypt was added subsequently. FLAG EA in its 23rd year is the grand old man and that says all about its relevance or rather retirement.

GBI/MENA hasn’t had trailblazing track record either. GBI owns the cable system, except for India-Gulf portion which is owned by Airtel. GBI & MENA (acquired by Telecom Egypt) went for complimentary FP swap. Thus, 1FP on GBI from the BU near Oman to Mumbai is owned by MENA or Telecom Egypt, with GBI owning 1FP on MENA cable from Oman to Italy.

These cable systems reek of stitched-up architecture that doesn’t evoke confidence on cutting edge upgrades to address hyperscale requirements and hence commercial longevity, other than captive inhouse requirements. Hence, they are expected to join the ranks of antiquated consortium cables.  That aside 1FP in new cable systems under development will have more capacity (T/FP) than the three cable systems put together and will be competitively edged to obsolescence unceremoniously, in the next 4-5 years. Isn't that long enough to make hay while the sun shines!

Middle East & North Africa

All the Mumbai-Marseilles submarine cables have branches to Middle East and incumbent carriers like Omantel, Etisalat and STC as commanding consortium members with portfolio of 10-14 submarine cable landings. Some go till UAE like IMEWE and SMW4. EIG goes a step further and lands in Oman and UAE. AAE-1 ventures deeper in the Persian Sea and lands in Oman, UAE and Qatar. TGN-Gulf branching from TGN-EA near Oman goes even deeper and lands in Bahrain and Saudi Arabia in addition to Oman, UAE and Qatar. GBI has the deepest coverage of Persian Sea and lands in Kuwait, Iraq & Iran in addition to Oman, UAE, Qatar, Bahrain and Saudi Arabia.  

The key submarine cable stopovers in North Africa are Djibouti and Egypt. Exceptions are SMW4 that lands in Tunisia and Algeria and SMW3 that lands in Morocco. However, Tunisia, Algeria and Morocco have had greater affinity to connect with France, Spain and Italy as evidenced by short span cable systems in the Mediterranean Sea. Djibouti is well known interconnect hub and lands 8 subsea cables. It includes SEACOM and EASSy that connect East African countries and leads to interesting cable hopping connectivity solutions. The role of Egypt does not merit any explanation with respect to submarine cables.

A key attribute of Mumbai-Marseilles subsea route has been Egypt Crossing or Red-Med Corridor, a terrestrial span of nearly 400 kms between Abu Talat and Zafarana or Suez and Alexandria. Incumbent monopoly of Telecom Egypt and their commercials for Egypt Crossing has been major contributor to submarine cable business case. Some industry observers have termed it extortionist and detrimental to development of new submarine cables. At the same time there hasn’t been any submarine cable that failed to take off due to commercials of Egypt Crossing.

Lukewarm discussion on Egypt bypass has been ongoing for last few years. Some consider “Egypt bypass” as cuss word and prefer to refer to is as diverse route. No one wants to burn bridges with Telecom Egypt and that’s well understood. However, a feasible solution through Israel, Jordan and Saudi Arabia has been challenging because of geopolitical landscape. However new IIPs have taken up the challenge, orchestrated harmonious collaboration, forged relationships for cross-border connectivity, convinced investors and developed end-to-end solution, and reportedly moving ahead with deployment. It remains to be seen when the first cable comes up with CIF announcement. .  

Cable Landing: Mumbai & Marseilles

Cable Landing will be a mixed experience in countries being considered. Mumbai and Marseilles will provide open cable landing and flexible architecture for fiber pair breakout, with the Landing Party not having any financial stake in the cable system. This disaggregation is important for open cable landing to be truly open. In other countries it would be unavoidable trade-off to join hands with incumbent monopolies. The belief that incumbents will also change, for incumbency no longer guarantees survivability going forward, does hold good for carriers in Middle East and North Africa, though the pace of change may be slower than expected. That said, their persistent appetite for investment in submarine cables remains truly praiseworthy.

Mumbai is the vantage point of interconnect for Mumbai-Marseilles and Mumbai-Chennai-Singapore cable systems. IMEWE-SMW4 has been well accepted combination and scored over EIG-<Mumbai-Chennai Terrestrial>-i2i. It has CLS from Tata Communications, Airtel, Reliance Jio, RCOM, Vodafone Idea and Sify. One other provider is reportedly working to join the CLS owners club with new CLS. That would mean eight (8) CLS in Mumbai. However not all are equally aligned to provide open cable landing. Not all CLS are equally suited for landing new submarine cables as well. Cable landing in some CLS will require Cable Crossing Agreement (CCA) with 6-8 submarine cables and oil & gas pipelines, and that would be a formidable challenge to navigate. Some might negotiate for opportunistic compensation like grant of capacity for CCA signoff. Some CLS by virtue of their location provide clean subsea route for cable landing that does not cross over existing submarine cables. It is advisable for new submarine cables to select CLS and Landing Party judiciously that provide open cable landing and carrier neutral FP breakout in letter and spirit.

With cable landing in seashore BMH, the location of PFE will determine the cable landing architecture. <BMH+PFE>-<DF BH>-<SLTE> or <BMH>-<DF BH>-<PFE+SLTE> will make big difference. The selection will be based on distance between the Datacenter hosing SLTE and BMH location. Usually beyond 5-7 kms, it is preferred to house the PFE adjacent to the BMH. Hence <BMH+PFE>-<DF BH>-<SLTE> is expected to be the preferred cable landing architecture. Just in case, <DF BH> unmistakably abbreviates dark fiber backhaul. This architecture calls for diversely routed 16FP dark fiber connectivity from the CLS i.e. <BMH+PFE> location to one or even multiple Datacenters if that’s preferred by individual FP owners. It is expected to prove challenging, first in terms of high count of dark fiber pair availability and second, in terms of dB loss per kilometre that is low enough for submarine optical transmission. Its likely to require fresh fiber to be laid or at best a combination of fresh fiber laying and dark fiber lease. It’s also likely to require optical line protection (OLP) solution. In all the dark fiber backhaul in Mumbai must be carefully designed, deployed and maintained.

Cable landing in Marseilles is majorly controlled by Interxion, followed by Orange. An unexpected newcomer, Omantel, also owns and operates CLS to land AAE-1. Marseilles famed for its streamlined approach to facilitate cable landing and pan European long-haul connectivity with multiple dark fiber provider options for Paris, London, Dublin, Frankfurt and Amsterdam and to CLS in UK and Ireland where Transatlantic cables land. It also has dark fiber connectivity to Lisbon to access to West African submarine cables WACSACE and Equiano (RFS 2021) and EllaLink (RFS 2020) from Brazil that land in Sesimbra, Seixal and Sines. However, 1700 kms of terrestrial connectivity is neither economical nor operationally efficient. That’s where the need for a purpose build submarine cable from Marseilles to Lisbon through Gulf of Gibraltar has been felt.

New Cable Systems

Three (3) new cables systems are expected and justifiably required for three-route-diversity. The first announcement came from Cinturion Corp in May 2019 for TEAS.  Originating in Milan/Genoa and Marseille and terminating in Mumbai, the  9,401 kms cable system will be routed through Gulf of Aqaba, Red Sea, Gulf of Aden, and Arab Gulf with two extensions landing in Mumbai. That’s quite an interesting architecture. After 8 months the second announcement comes from Reliance Jio for IEX, along with IAX for Mumbai-Chennai-Singapore. The 3rd cable system is expected to be SMW6 and its announcement is expected to take 3-6 months. Note that these are merely announcements, it does not confirm CIF or disclose nxFP and T/FP. That said TEAS, IEX & SMW6 is well balanced with MIST, IAX & SMW6. On second thought however it would not be fair to include SMW6, IAX and IEX with TEAS and MIST. While IAX and IEX land with Reliance Jio and to start with it will own all the 12/16 FP, SMW6 reportedly lands with Airtel in Mumbai and Chennai and it will own 1FP. How far will they be open cable system remains apprehensively questionable.  

The 3rd open cable system to land in Mumbai, excluding IEX and SMW6, is expected to take 3-6 months to reveal their identity. Considering these will be 16FP systems with no less than 18T/FP, anything beyond the 3rd would be stretch on optimism and supply-demand practicality, if not outrightly irrational. Investor scepticism can’t be ignored either.

OTTs, IIPs and Telco Carriers are expected to collaboratively orchestrate the new submarine initiatives. It however would not be true to say that Telco Carrier involvement is largely relegated to provide open cable landing. Not at least for Middle East where incumbent Telco Carriers expected to be major co-investor in some of the new submarine cables.

New submarine cables are likely to focus on Mumbai-Marseilles express route. Provision of branches to Middle East could be planned for Phase-2, unless someone like Omantel, Etisalat or Du signs up as co-investor or anchor tenant. It also merits consideration that Middle East countries are still Europe oriented and that’s reflected in asymmetric growth in IPT. This trend could be offset by Microsoft and AWS setting up Datacenters in UAE and Bahrain and create traction for Marseilles-Oman/UAE-Mumbai connectivity. Thus, a Middle East branch landing in Oman or UAE on day one can’t be ruled out.

How challenging will it be for the new cable systems to sign up anchor customers, usual suspects being the big-4 OTTs. It certainly will be uphill task, particularly with new untested cable routes with terrestrial spans. Commercial and contractual safeguards may be offered to assuage anchor customers concerns. Others may choose to start with capacity lease to experience the cable system performance before deciding on fiber pair acquisition. However, growth of cloud and content services in India will make continued nx100G acquisitions nonviable as its already happening with metro and long-haul connectivity. Hence requirement of long-term solution will drive risk-and-reward play with fiber-pair-ownership.

An interesting observation is, if one of the new submarine cables or the first one to reveal its identity is 16FP it will lead to fiber pair economics of ~$35M/FP and that will compel other cable systems to go with 16FP design for competitive equivalence. 

Does three new submarine cables make Mumbai-Marseilles a highly commoditized subsea route that is set to witness severe price drop as the cable systems reach RFS in 2022-23 and customers exhibit growing reluctance to go for IRU deals with front-end payments! Isn’t that a generic statement that holds good for any subsea route. But it has industry favourite rebuttal – reliable connectivity demands three diverse routes and hence at least three subsea cables are needed. Thus three (3) cable systems does not necessarily mean 3X capacity or FP. The arguments can continue, but the recent episode of Seabras-1 filing voluntarily for restructuring under Chapter 11, that came to light in the last week of December 2019 as fallout of downward spiral of market price decline with BRUSA and MONET, will remain a grim remainder of competitive brutality and its consequence.

Let’s Conclude

New Mumbai-Marseilles submarine cables in the making present an interesting combination of subsea and terrestrial routes with undertones of geopolitical dynamics, direct and branch cable landing in countries with incumbent monopoly or duopoly at best. It will require thoughtful route planning, harmonious collaborations, large investment and take longer for deployment. It will increase the count of submarine cables landing in Mumbai, including that of Mumbai-Chennai-Singapore, from 10 to 16 in the next 3-4 years and move ahead of Marseilles (5 cable landing at present) as new global hub of open cable landing. Datacenters with SLTE hosted will bring path-breaking disruption to access subsea connectivity. Let's hope the protection of Poseidon prevails and drives the new submarine cables to successful deployment.