DXN (ASX: DXN) backlog hits A$40.9m as AI data-centre contracts reshape FY27 outlook

DXN Limited (ASX: DXN) has entered FY27 with its largest contracted backlog on record after artificial-intelligence infrastructure contracts sharply changed the scale of its forward workload. The company’s August 31 FY26 update showed backlog rising from A$23.5 million at June 30 to A$40.9 million by August 30 following new contract awards, while management expects roughly 45% of the backlog to convert into revenue during the first half of FY27. The Australian Securities Exchange classified the August 31 release as price-sensitive and published it before the market opened.

That forward position contrasts sharply with the year just completed. FY26 revenue declined 33% to A$10.1 million from A$15.1 million as customer-side project deferrals affected the first half, while gross profit fell to A$1.8 million and the loss attributable to shareholders widened to A$6.6 million. Underlying EBITDA from continuing operations was negative A$3.7 million, compared with positive A$0.5 million in FY25.

The resulting comparison is unusually stark. DXN’s A$40.9 million August backlog is about 4.05 times FY26 revenue, while the approximately 45% expected to convert during the first half mathematically represents around A$18.4 million of potential revenue if projects progress according to management’s delivery assumptions. That single six-month conversion target would therefore exceed the company’s entire FY26 revenue by roughly 82%.

How did artificial-intelligence infrastructure change DXN’s backlog so quickly?

DXN’s strategic inflection began with an A$8.8 million contract signed in June for a 1.36MW artificial-intelligence high-performance-computing modular data centre for a publicly listed United States neocloud operator. The module is being manufactured at DXN’s Welshpool facility and is expected to be commissioned at the customer’s mainland United States site within approximately six months. Management has linked successful completion of that proof-of-concept deployment to an indicative campus-scale follow-on opportunity valued at more than US$200 million, although that larger opportunity is not presently contracted revenue.

A second AI infrastructure contract followed in August, worth A$12.2 million for a turnkey 2MW modular data centre. DXN said the customer required a deployment timetable that conventional construction could not satisfy, reinforcing the company’s core argument that factory-built modular infrastructure can address the unusually fast capacity requirements of artificial-intelligence compute operators. The company has also secured work involving Melbourne Airport and the Solomon Islands Submarine Cable Company since year-end.

Together, the two disclosed AI contracts are worth roughly A$21 million, more than twice DXN’s entire FY26 revenue. They do not necessarily translate into revenue in a single reporting period, but they explain why the current backlog provides a much stronger forward indicator than the historical income statement.

DXN’s broader identified pipeline contains 99 projects, with approximately 21% linked to global AI compute infrastructure customers. Seven were described as verbal wins or in contracting and another eight were in final negotiations as of the company’s pipeline snapshot, creating additional potential conversion beyond the current contracted backlog.

Why is the US$200 million follow-on opportunity important but still highly conditional?

An indicative US$200 million opportunity would be transformative relative to DXN’s existing scale. At the August 28 pre-announcement market capitalisation of approximately A$136.4 million, even a portion of a project of that size would substantially change the company’s revenue base. DXN shares had already increased sharply during August, including a 41.8% gain on August 24 following the second AI contract, and closed August 28 at A$0.37 compared with A$0.28 one week earlier.

The larger opportunity nevertheless remains contingent on successful delivery and commissioning of the initial 1.36MW proof-of-concept unit. It should therefore be viewed as an addressable follow-on opportunity rather than backlog, an awarded contract or forecast revenue. That distinction is especially important after DXN’s equity value increased more than sixfold year on year, according to market-cap data through August 28.

Execution risk is also becoming more consequential because DXN is expanding manufacturing while delivering increasingly sophisticated high-density infrastructure. Its AI modules are designed to support rack densities of up to 150kW and extensive direct-to-chip liquid cooling, while factory fabrication is intended to reduce deployment time relative to conventional data-centre construction. Those attributes create commercial differentiation only if DXN can deliver schedule, quality and commissioning performance as order sizes increase.

Does DXN have enough capital to convert the A$40.9 million backlog?

DXN finished FY26 with approximately A$11 million of cash, compared with A$3.1 million a year earlier, after completing a A$7 million institutional capital raising during June. Contract liabilities increased to A$3.8 million as customer deposits flowed into recently awarded projects, while inventories and work in progress increased as manufacturing activity accelerated.

The company is simultaneously expanding its production footprint. A wholly owned manufacturing facility in Johor, Malaysia, and a proposed east-coast Australian facility are both targeted to begin production during the second quarter of FY27. The capacity expansion is intended to position DXN for the larger modular data-centre pipeline, but it creates a working-capital and execution challenge because manufacturing capacity must be added before some potential future projects become contracted revenue.

The A$11 million cash balance equals about 27% of the current A$40.9 million backlog, although that comparison should not be interpreted as the cash required to deliver the projects because customers provide deposits and revenue is collected through contract milestones. It nevertheless demonstrates why cash conversion, progress payments and disciplined manufacturing expansion matter as the company’s project scale increases.

What is the correct market-reaction reference for DXN’s August 31 release?

DXN closed at A$0.37 on Friday, August 28, down 5.1% in that session but still approximately 32% above its August 21 close of A$0.28. Its 52-week range through August 28 was A$0.016 to A$0.44, illustrating how dramatically market expectations had already changed before the FY26 release.

The FY26 results announcement was released at 8:38 a.m. Australian Eastern Standard Time on Monday, August 31. Because the August 31 trading session had not finished when this article was prepared, the August 28 close should be treated as the pre-announcement reference price rather than describing an incomplete Monday move as the definitive market reaction.

DXN now has to convert a compelling forward-order story into financial performance. The company enters FY27 with backlog roughly four times FY26 revenue, a stronger cash position and two meaningful AI infrastructure contracts, but it is starting from a year in which revenue declined and losses widened. Delivering the first AI module on schedule could therefore become more important than another headline pipeline number because it is the gateway to the much larger customer opportunity management has identified.