SUBCO unveiled a landmark project in transpacific connectivity: the APX East cable. This new hypercable will establish a first-of-its-kind, direct optical link between Sydney, Australia, and San Diego, California, USA, with no intermediate landings. 

Designed specifically to meet the unprecedented bandwidth demands of Australia’s burgeoning artificial intelligence (AI) industry, the APX East project is slated for completion in Q4 2028 with an investment in the ballpark of US$500 million (AU$747 million)

The announcement directly addresses a critical infrastructure gap. As AI factories requiring immense computational power are planned for Australia, the limiting factor is increasingly seen as international data capacity rather than domestic power or data centres.

"With all the talk of AI factories, people are forgetting that the longest lead item for Australia isn't going to be power, land, data centres or chips, it's going to be international connectivity at AI scale," stated Bevan Slattery, Founder & Co-CEO of SUBCO. "APX East will be a critical enabler for Australia’s aspirations to become a leader in the AI world."

APX East
APX East, Image courtesy: SUBCO

 

A Technical Leap in Cable Design

APX East is engineered as a pure "express" route, a significant technological advancement over existing systems. Its key innovations include:

  • Direct Optical Path: SUBCO stated that "APX East will be first express cable system between Australia and the United States not requiring optical regeneration1, capable of being single-end power fed across the entire system" and "APX East will be the single longest continuous subsea optical path in the world2, and the longest system able to support single-end power."

  • Simplified Operations: The entire system can be powered from a single end during a fault condition. For hyperscale customers, this means they only need to install terminal equipment (SLTE) in Sydney and California, with no need for complex infrastructure in between.

  • Enhanced Security & Latency: The all-deepwater route avoids shallow coastal zones, improving physical security and offering the lowest latency path between the two nations.

Strategic Importance for AI and Sovereignty

The project is a direct response to a quantified need. Hyperscalers and neoclouds are looking to deploy 3GW of AI factories in Australia between now and 2028. SUBCO estimates that planned AI factories in Australia will require between 75 to 150 Terabits per second (Tbps) of international capacity to function. APX East, with its 16 fibre-pair design, is being built on an accelerated timeline to meet this 2028 demand window, arguing that systems with later completion dates will be too late.

Beyond capacity, the cable carries strategic national significance. SUBCO highlights it as "Australia’s first sovereign-owned international hypercable," a move designed to reduce the nation's reliance on infrastructure owned by US-based technology hyperscalers for its critical data exports, particularly the output of future AI services.

Project Timeline and Future Expansion

The primary Sydney-California trunk is firmly scheduled for Ready for Service (RFS) in Q4 2028. The project plan also includes optional future branches to Hawaii (Q4 2029) and Fiji, which are additive and not required for the main system's operation. These would provide additional route diversity and resilience.

By taking a new, diverse landing site north of Sydney's existing cable protection zone, APX East also mitigates terrestrial congestion and risk. This ambitious project positions SUBCO to provide the foundational digital infrastructure it believes is essential for Australia's next phase of economic and technological development.

SUBCO's Subsea Cable Portfolio

In addition to the newly announced APX East, SUBCO owns the following subsea cable systems:

  • SMAP: An Australian transcontinental system connecting Sydney, Melbourne, Adelaide, and Perth, which is expected to be ready for service in Q1 2026.
  • Oman Australia Cable (OAC): A 9,800 km cable connecting Perth, Australia, to Muscat, Oman, which launched for commercial service in November 2022.
  • INDIGO-West: Part of the Indigo system, this cable connects Perth to Singapore via Jakarta. SUBCO has increased its spectrum investment on this system.
  • INDIGO-Central: This system provides connectivity between Perth and Sydney. SUBCO is the equal largest capacity owner on this system.

Scrutinizing the Technical Claims

SUBCO’s press release makes two bold technical claims that warrant closer examination from an industry perspective.

1. On "Not Requiring Optical Regeneration"
SUBCO's claim that APX East will be a system "not requiring optical regeneration" is a nuanced statement that could be considered technically misleading. In subsea cable engineering, "regeneration" can refer specifically to Optical-Electrical-Optical (O-E-O) conversion at intermediate landing points (like Guam or Hawaii), which fully reshapes the signal. The APX East system indeed appears to avoid this by having no intermediate landings.

However, what is definitively required for any subsea cable of this length are submerged optical repeaters (often erbium-doped fiber amplifiers, or EDFAs) spaced along the seabed cable to boost the optical signal. Therefore, a more precise description is that APX East eliminates intermediate-point O-E-O regeneration but will still rely on a chain of in-line optical amplification. The claim highlights a design that simplifies network architecture by removing intermediate Points of Presence (PoPs), but it does not mean the signal travels unaided.

2. On "The Single Longest Continuous Subsea Optical Path"
The claim that APX East will be "the single longest continuous subsea optical path in the world" is currently unverifiable and skeptical. SUBCO has not publicly announced the exact distance of the Sydney-California express route. Industry estimates suggest a direct great-circle path would be approximately 12,000 to 13,000 km.

Crucially, there are existing transpacific systems that also employ continuous optical paths over similar distances without electrical regeneration at intermediate points. Without SUBCO disclosing the precise engineered path length and the specific technical parameters defining a "continuous optical path," this claim appears to be a marketing assertion of a "first" rather than a proven, singular engineering achievement. It may be intended to emphasize the system's ambitious single-span design rather than to denote a literal, unrivaled global record.