Firmus Raises $2 Billion for APAC AI Infrastructure
Firmus confirmed a $2 billion equity round on August 7, 2026, valuing the AI infrastructure company at more than $10.5 billion. Coatue and NVIDIA made follow-on investments, while Blackstone-managed funds and Jane Street also participated; the company said the capital will support Project Southgate in Australia and preparations for wider Asia-Pacific expansion.
Firmus confirmed a $2 billion equity round on August 7, 2026, taking its post-money valuation above $10.5 billion. Capital Brief reported follow-on participation from Coatue and NVIDIA, with new investment from funds managed by Blackstone Tactical Opportunities and other Blackstone vehicles, plus Jane Street.
Startup Daily reported the same transaction as A$2.85 billion at an A$15 billion valuation. The Australian-dollar figures describe the same round rather than a separate financing event.
Capital for Project Southgate
Capital Brief reported that Firmus plans to use the investment to accelerate the next phase of Project Southgate, its Australian AI factory rollout, while preparing for expansion into other Asia-Pacific markets. Firmus co-CEO Oliver Curtis said the company intends to scale across Australia and advance early work connected to its recently announced development in Indonesia.
Firmus builds training and inference infrastructure around NVIDIA's DSX AI Factory reference architecture. The financing therefore joins capital, accelerator supply, data-center construction, networking, cooling, and power delivery in a single expansion program. It does not by itself establish when new capacity will become generally available to external customers.
What the round changes
The new valuation is almost twice the $5.5 billion post-money valuation Firmus announced with its April 2026 financing. That change shows how strongly private investors are pricing prospective demand for AI compute, but valuation growth is not evidence that the planned facilities are already operational.
For ML teams, the practical effect depends on execution. New regional capacity can improve access to GPU-backed services and reduce distance to users and governed data, while actual workload decisions still depend on service availability, latency, data-residency terms, interconnects, reliability commitments, and price. Those operating details remain more consequential than the headline funding amount for teams choosing where to train or serve models.
Key Points
- 1Firmus confirmed a $2 billion equity round at a post-money valuation above $10.5 billion.
- 2The company said the capital will accelerate Project Southgate in Australia and preparations for wider Asia-Pacific expansion.
- 3For practitioners, usable regional GPU capacity will depend on delivery, service terms, networking, power, and reliability rather than the financing headline alone.
Scoring Rationale
The $2 billion round materially increases the capital available for a large AI infrastructure rollout and nearly doubles Firmus' reported valuation. Its direct effect on practitioners remains dependent on delivery and commercial availability of the planned compute capacity.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
