The Numbers
Firmus is spending $4.39 billion, or 61.4% of its $7.1 billion ASX raise, on servers and deployment infrastructure. Break that down: $3.46 billion for Nvidia hardware, $929 million for deployment infrastructure. Another $2.21 billion goes to liquidity reserves and financing costs. The rest covers working capital and IPO costs.
Shares are priced at $11, targeting a $43.7 billion market cap. IPO date is October 23, code AIF. Prospectus drops October 12.
What Firmus Actually Does
This is not a startup anymore. Firmus is a capital-heavy AI infrastructure vendor operating data centres across Australia and Southeast Asia: Tasmania, Melbourne, Canberra, Sydney, Perth, plus Singapore, Indonesia and Malaysia. They call them "AI factories." The tech is rack-scale Nvidia GPU systems with proprietary cooling.
The company started in 2019 doing Bitcoin mining in Tasmania. Co-CEOs Oliver Curtis and Tim Rosenfield pivoted to AI compute. Now Firmus is being called Australia's third-largest data-centre company by valuation.
The Funding Stack
Firmus raised $500 million in November 2025, another $280 million in June 2026, then closed a $2 billion strategic equity round. Year-to-date equity: $2.5 billion. Backers include Nvidia, Blackstone, Coatue and Jane Street. There is also a Blackstone-led $10 billion debt facility.
Nvidia sits on both sides here: they backed Firmus with $330 million in September 2025, and now Firmus is spending billions on Nvidia hardware. They also signed a multi-year compute deal with OpenAI for two Malaysian sites.
What This Means for Sales Teams
If you are selling into infrastructure or enterprise AI buyers, Firmus is now a major procurement story. They are deploying compute at scale across ANZ and Southeast Asia. Customer base skews large enterprise and AI platform buyers.
LinkedIn lists them at 201-500 employees, headquartered in St Leonards, Tasmania. No public hiring numbers yet post-IPO, but scaling infrastructure deployments across six countries suggests GTM expansion is coming. Watch for enterprise AE and AM roles targeting AI workload buyers.
The Nvidia hardware spend also signals where compute capacity is moving: this is hyperscaler-adjacent AI infrastructure, not traditional cloud. If your patch includes AI buyers or infrastructure procurement, this is your market shifting.