SharonAI Holdings announced a six-year strategic compute collaboration with NVIDIA that will support the deployment of a 72MW AI factory infrastructure footprint in Australia. The agreement enables Sharon AI to scale its cloud AI platform with up to 40,000 NVIDIA Grace Blackwell GB300 GPUs, targeting demand from enterprises, AI startups, universities, government agencies, and research institutions seeking sovereign AI compute capacity.
The collaboration will leverage NVIDIA’s DSX AI factory architecture and introduces a revenue-sharing model designed to accelerate deployment of large-scale AI infrastructure. Under the arrangement, Sharon AI will provide NVIDIA-powered cloud services while NVIDIA receives traditional hardware revenue along with a share of cloud revenue generated by supported capacity. The companies said the structure reduces the capital burden typically associated with AI infrastructure expansion while increasing access to advanced AI compute resources.
Following the agreement, Sharon AI’s total AI factory capacity rises to 132MW, with 102MW already contracted to customers. The company expects to deploy more than 55,000 NVIDIA GPUs across its infrastructure footprint by mid-2027. Sharon AI, which operates as a certified NVIDIA Cloud Partner, said the additional capacity strengthens its position in Australia’s emerging sovereign AI ecosystem and supports the build-out of AI factories serving commercial and public-sector organizations.
• Collaboration supports deployment of 72MW of new AI data center capacity in Australia
• Up to 40,000 NVIDIA Grace Blackwell GB300 GPUs planned under the agreement
• Sharon AI’s total AI factory capacity expands to 132MW
• 102MW of capacity already contracted to customers
• Company expects more than 55,000 NVIDIA GPUs deployed by mid-2027
• Revenue-sharing structure aligns cloud service growth with infrastructure investment
• Deployment based on NVIDIA’s DSX AI factory design
“This strategic compute collaboration with NVIDIA marks a pivotal moment in Sharon AI’s mission to deliver sovereign, large-scale AI compute infrastructure. Securing access to 72MW of data center capacity enables us to deploy up to an additional 40,000 Grace Blackwell GB300 GPUs, providing access to accelerated compute to enterprise, startup and AI native customers who otherwise may not have been able to access it,” said James Manning, Co-Founder and Chief Executive Officer of Sharon AI.
🌐 Analysis: The agreement reflects a growing trend in which NVIDIA works more closely with cloud service providers and emerging “neocloud” operators to accelerate deployment of AI infrastructure. Rather than relying solely on hyperscale cloud providers, NVIDIA increasingly supports regional AI factory projects designed to provide sovereign AI capacity for governments, enterprises, universities, and startups.
🌐 Analysis: The scale of the planned deployment is notable. A fully populated 40,000-GPU GB300 environment would rank among the largest AI infrastructure projects announced in the Asia-Pacific region. The revenue-sharing structure also signals evolving business models for AI infrastructure, as vendors seek recurring participation in cloud economics rather than relying exclusively on hardware sales.
| Company Profile: SharonAI Holdings Inc. (NASDAQ: SHAZ) | |
|---|---|
| Headquarters | New York, USA and Australia (Operates specialized high-performance cloud infrastructure across APAC) |
| Leadership | James Manning (Co-Founder & CEO) | Andrew Penn (Chairman of the Board, appointed May 2026) |
| Capital Profile & Financing | • Nasdaq IPO (Feb 2026): Raised gross proceeds of US$125 million. • Convertible Senior Notes (May 2026): Closed private offering of US$350 million in aggregate gross proceeds. • Debt & Credit Facilities: Secured access to a debt facility of up to US$500 million from USD.AI, alongside an infrastructure revenue-share facility of up to US$200 million via Digital Alpha. |
| Industry Category | Neocloud Provider / Accelerated High-Performance Computing (HPC) Infrastructure |
| NVIDIA Relationship | Certified NVIDIA Cloud Partner. Operating under a six-year strategic compute collaboration using a revenue-sharing and credit-support model designed to align infrastructure deployment with cloud-service growth. |
| AI Factory Power Capacity | • Total Footprint: 132 Megawatts (Expanded via a 72MW June 2026 collaboration agreement to enable new capacity in Australia). • Contracted Capacity: 102 Megawatts pre-sold to end enterprise and cloud customers. |
| Hardware Deployment Target | Plans to scale up to an additional 40,000 NVIDIA Grace Blackwell GB300 GPUs utilizing NVIDIA’s DSX AI factory design, aiming for a total macro target of 55,000+ active GPUs deployed by mid-2027. |
| Strategic Infrastructure Ecosystem | Cisco (Go-to-market secure AI factory partnership using UCS architecture), NEXTDC (Co-location data center infrastructure), and World Wide Technology (WWT). |
| Key Asset Realignment | Completed the early sale of its 50% holding in the Texas Critical Data Centers (TCDC) joint venture for US$70 million to recycle capital into its core Australian and Asia-Pacific GPU cloud business. |
