Summary

  • Light Source Communications says a second hyperscale tenant has signed for its roughly 500-mile Tulsa–St. Louis dark-fibre route, up from the one tenant disclosed in February.
  • The new signature can reduce dependence on a single anchor customer, but LSC has not disclosed either tenant, the capacity or route segment contracted, value, term, service date, acceptance conditions or billing.
  • Engineering is described as substantially complete and early construction as under way; the route’s forecast completion remains the third quarter of 2027. The next receipts are physical delivery, end-to-end testing, customer acceptance and billable service—not another tenant count.

Two signatures are not two operating circuits

In February, LSC introduced a long-haul dark-fibre project linking data centres in St. Louis and Tulsa. It described a 500-mile underground route, eight in-line amplifier sites and one hyperscale tenant, with completion anticipated in the third quarter of 2027. On 1 September, the company said it had signed a second hyperscale tenant. Engineering was then “substantially complete” and initial construction phases were under way.

That is a meaningful change in the demand record. It is not a change in the route’s operating state. The September announcement does not say that a strand has been handed over, that an amplifier site has been energised, that an optical-loss budget has passed, or that either customer has accepted service. It preserves the Q3 2027 construction forecast.

The distinction matters because long-haul fibre converts through several separate clocks. A commercial agreement can support a build decision. Civil works create the physical path. Fibre placement and splicing create continuity. Amplifier huts and points of presence must be ready. End-to-end tests establish whether the promised route works. Contractual acceptance can then start a service or billing clock. A tenant may still need to install and operate its own optical equipment before traffic moves over dark fibre.

LSC has disclosed the first clock twice. The later clocks remain ahead.

The second tenant changes concentration before utilisation

One anchor customer makes a corridor legible enough to build, but it can also concentrate the project’s commercial outcome. A second independent tenant can spread fixed civil costs and reduce reliance on one renewal, one deployment timetable and one credit decision. That is the strongest economic inference supported by the announcement.

The size of the improvement is unknowable from tenant count alone. LSC does not say whether both customers lease fibres end to end, take different segments, require different points of presence or commit to the same start date. It does not publish fibre-pair counts, contract values, prices, terms, deposits, minimum commitments or cancellation rights. A second signature could represent comparable demand, a much smaller commitment, or capacity whose activation follows a different timetable.

The correct denominator is committed physical scope and economics relative to the route that is ultimately built and accepted. “Two tenants” supplies only the numerator’s headcount, not its weight. It cannot be converted into doubled utilisation, revenue, cash flow or project return.

That boundary is especially important when AI demand is used to frame infrastructure announcements. A hyperscale label says something about the class of customer. It does not reveal how many strands the customer has reserved, what service level it requires or whether a particular AI cluster will use the corridor.

Roughly 500 miles still have a delivery ledger

LSC calls the route 500 miles in February and more than 500 miles in September. Its network page says the corridor will connect data centres in the two cities, link with its 130-mile Tulsa metro ring and add further points of presence. The design is entirely underground and includes eight in-line amplifier sites intended to preserve signal quality.

Those specifications establish scope, not completion. “Substantially complete” engineering is not a percentage of route miles built. “Initial construction phases” does not identify cleared right of way, conduit installed, fibre pulled, splices completed, huts powered or tests passed. The sources provide no permit register, construction schedule by segment or acceptance record.

A useful project ledger would therefore separate at least six measures: engineered miles, accessible or permitted miles, conduit-complete miles, fibre-and-splice-complete miles, tested miles and customer-accepted miles. Amplifier and point-of-presence readiness need their own counts. A single completion date compresses all of these dependencies into one forecast.

The fact that the route is intended to be underground can improve protection from some surface hazards, but it does not eliminate construction, relocation, access or repair risk. Nor do eight planned amplifier sites prove eight working sites. The financial value appears only when the physical and contractual ledgers meet.

The revenue clock is deliberately blank

Neither announcement states when billing begins. There is no disclosed prepayment, installation charge, recurring fee or accounting policy. It would therefore be wrong to treat the second signing as recognised revenue or even as a disclosed backlog amount.

For a wholesale fibre owner, the order of evidence matters. A signed contract may improve financing credibility. A completed path may allow testing. Acceptance may unlock billing. Invoices may create receivables. Collection creates cash. The timing and risk of each step depend on terms that are not public here.

This is not a reason to dismiss the event. It is a reason to value it correctly. LSC has moved from one disclosed demand source to two while construction remains early. The signature reduces one form of uncertainty and leaves the delivery and monetisation uncertainties intact.

What would change the assessment

The most informative next disclosure would not be another adjective about AI demand. It would be a bridge from customers to capacity and from capacity to delivery: contracted route scope, fibre pairs or equivalent capacity, segment coverage, milestone dates, acceptance criteria, ready-for-service notices and the start of billing. Aggregated data could preserve customer confidentiality.

Evidence that the two contracts cover substantial, non-overlapping or independently valuable capacity would strengthen the diversification case. Evidence of route completion, optical testing and customer acceptance would move the story from demand to delivery. Disclosed recurring revenue or collections would finally connect the corridor to realised economics.

Until then, the September signature deserves a precise label: a second demand receipt on a still-unfinished route.

Sources