Sharon AI Pursues ASX Listing After Firmus IPO Withdrawal

Australian AI infrastructure provider Sharon AI is pursuing a potential listing on the Australian Securities Exchange (ASX), seeking to give domestic investors access to a business that already trades on Nasdaq. The announcement comes after fellow Australian data-centre company Firmus withdrew its planned initial public offering, putting renewed attention on the valuations and funding requirements of large-scale artificial intelligence infrastructure projects.

Sharon AI co-founder and chief executive James Manning said the company wanted to make its shares more accessible to Australian retail investors, according to reporting published on October 12. The company has not announced a definitive timetable for an ASX listing. Its existing Nasdaq shares trade under the ticker SHAZ, following a US initial public offering priced at US$30 a share in February 2026.

The proposed Australian listing would therefore be an additional market presence rather than Sharon AI’s first public-market debut. Its timing will be watched closely as investors reassess how much they are willing to pay for companies building the computing capacity needed to run AI applications.

Firmus withdrawal puts AI valuations under scrutiny

Firmus’s planned ASX float was withdrawn on October 9 after the company struggled to secure sufficient investor support for an offering targeting a valuation of approximately A$44 billion. The proposed transaction had been expected to raise about A$5 billion, making it one of the most significant listings contemplated for the Australian market in recent years.

The withdrawal followed concerns about valuation, execution risk and the scale of infrastructure that still needed to be delivered. Firmus had ambitious plans for AI-focused data centres, but investors were being asked to assess those plans against the company’s existing operating capacity and substantial future capital requirements. The company has since turned to alternative funding options.

The episode does not establish that investors have lost confidence in AI infrastructure as a whole. Demand for computing capacity continues to support investment in data centres and specialised chips. Instead, the failed float illustrates the difficulty of translating expectations for future AI demand into a valuation that public-market investors are prepared to accept.

For companies seeking to raise capital, the distinction between projected capacity, signed contracts, operational infrastructure and cash actually received can become decisive. Large headline contract figures may indicate potential demand, but they do not automatically show how quickly a company can deliver services or earn a profit.

Sharon AI’s contracts and Nvidia partnership

Sharon AI operates as a specialised AI cloud infrastructure provider, often described as a “neocloud” company. Rather than relying exclusively on a fully owned data-centre network, its model involves securing infrastructure capacity and providing customers with access to advanced computing resources.

A major element of its strategy is a six-year collaboration with Nvidia, announced in June 2026. The agreement has a potential contract value of approximately US$4.88 billion and is intended to support 72 megawatts of AI data-centre capacity in Australia, using up to 40,000 Nvidia Grace Blackwell GB300 GPUs. The arrangement includes a revenue-sharing and credit-support structure.

The Nvidia agreement is significant because access to high-performance GPUs is a central requirement for many AI training and inference workloads. However, the stated contract value should not be read as cash already earned or guaranteed profit. Realising the agreement’s value depends on deploying the required infrastructure, meeting operational standards and delivering services in line with contractual requirements.

Sharon AI’s own regulatory disclosures identify risks associated with procurement, installation, testing, customer acceptance and ongoing infrastructure availability. These requirements are important because delays or performance shortfalls can affect the timing and economics of a large contract.

Billions in contract value, but a different revenue picture

In its second-quarter 2026 results, Sharon AI reported total contract value of approximately US$8.8 billion. It also reported quarterly revenue of US$1.9 million, up 412% year on year, and adjusted EBITDA of US$0.6 million. The company recorded a net loss of US$430.4 million, which included substantial non-cash accounting items, primarily a fair-value loss on convertible notes. It reported US$1.9 billion in cash and cash equivalents at June 30.

These figures need to be considered together. Total contract value reflects the stated value of agreements over their contractual periods; it is not the same as revenue recognized during a quarter. Revenue depends on services being delivered and recognized under the applicable accounting rules, while profitability also reflects operating costs, financing expenses and other accounting items.

The difference between contract value and current revenue is not, by itself, proof that contracts are unreliable. Long-term infrastructure businesses may spend heavily before new capacity becomes operational and revenue scales. But investors must evaluate the delivery schedule, financing requirements, customer concentration and the costs of bringing promised capacity online.

For Sharon AI, the question is how quickly its contracted demand can translate into recurring revenue and sustainable financial performance as its infrastructure expands.

What a potential ASX listing could mean

An Australian listing could broaden Sharon AI’s investor base and make its shares more accessible to local investors. It could also increase scrutiny from domestic institutions and retail shareholders, who would assess the company’s financial disclosures, capital needs and progress against its deployment commitments.

The timing remains uncertain. The Firmus episode may encourage investors to examine assumptions more carefully, but it does not establish the outcome of Sharon AI’s proposed listing. The two businesses have different structures and project profiles, and their respective valuations must be assessed on their own evidence.

A listing would also come with practical considerations, including the requirements of the ASX, the structure of any secondary listing, liquidity across markets and the costs of maintaining compliance. Sharon AI has not yet provided a definitive timetable in the reporting available on October 12.

What investors will watch next

The next milestones include any formal ASX timetable or listing documentation, updates on Sharon AI’s infrastructure deployment, and evidence that contracted computing capacity is turning into recognized revenue. Investors will also be watching cash requirements, financing arrangements and operating performance as the company expands.

Firmus’s withdrawal has sharpened the debate over how AI infrastructure companies should be valued. Sharon AI enters that debate with an existing Nasdaq listing and significant reported contracts, but the commercial test remains whether those commitments can be delivered profitably.

For Australia’s technology market, the potential listing could provide another opportunity to attract investment into AI infrastructure. Its significance will ultimately depend not on the size of headline contracts alone, but on execution, financial transparency and the company’s ability to demonstrate durable operating results.