PlatformsStrategic PartnershipSeptember 21, 2026

Sharon AI NVIDIA GPU deployment secures 8,200 Blackwell units in Sydney

Sharon AI NVIDIA GPU deployment secures 8,200 Blackwell units in Sydney

GreenSquare has signed Sharon AI as an anchor tenant for its SYD1 data center in Western Sydney, marking a significant deployment of 8,200 NVIDIA Blackwell Ultra GPUs. The facility will support sovereign AI infrastructure for model training and inference, with operations scheduled to begin in Q4 2026.

Key Takeaways

  • GreenSquare will deploy 8,200 NVIDIA Blackwell Ultra GPUs and HGX B300 systems at its SYD1 campus
  • The project reuses an existing data center building to reduce carbon footprint and accelerate delivery timelines
  • Sharon AI will use its AI Factory platform to provide local sovereign infrastructure for Australian customers
  • GreenSquare is currently developing a 1GW pipeline across New South Wales and Victoria backed by Partners Group

Why It Matters

This agreement establishes a commercial benchmark for high-density AI infrastructure in Australia, proving that brownfield data center redevelopments can meet the power requirements of next-generation hardware. By securing 8,200 NVIDIA Blackwell Ultra GPUs, Sharon AI is positioning itself as a primary local alternative to global hyperscalers for organizations requiring sovereign data processing. This move signals a shift toward specialized neocloud providers that use colocation to bypass the lengthy construction cycles of traditional greenfield builds. Watch for whether GreenSquare can convert its 1GW development pipeline into similar long-term contracts as regional demand for inference capacity scales through 2027.

Additional Context

GreenSquare's SYD1 campus is part of a broader wave of brownfield data center conversions targeting AI workloads in Australia. Sharon AI has positioned itself as a sovereign AI infrastructure provider backed by Partners Group capital, deploying NVIDIA HGX B300 systems that require liquid cooling and power densities exceeding 100kW per rack. The deal mirrors a pattern across Asia-Pacific where neocloud operators are securing anchor tenants before completing full facility buildouts to de-risk capital expenditure on GPU clusters that depreciate faster than traditional server hardware. On the business and procurement side, Sharon AI's commitment to 8,200 Blackwell Ultra units represents one of the largest single-site NVIDIA deployments in Australia outside hyperscaler facilities. Mux and Synamedia have partnered to integrate real-time QoE signals with CDN switching decisions, illustrating how video infrastructure buyers increasingly evaluate compute proximity and delivery cost together. For video streaming platforms considering sovereign inference workloads, the GreenSquare-Sharon AI arrangement offers a colocation model where GPU capacity sits in the same metro as content delivery points of presence, reducing round-trip latency for real-time AI tasks such as content moderation and automated dubbing. From a technical standpoint, the NVIDIA Blackwell Ultra architecture that Sharon AI is deploying delivers significant throughput gains for video AI inference compared to prior-generation Hopper chips. Bitmovin's 2026/2027 Video Developer Report found that 98 per cent of video professionals now use AI or ML in their workflows, with audio transcription, translation, and foreign dubbing cited as the top application at 48 per cent of respondents. Mux launched its Robots API in 2026, moving video AI analysis natively inside its platform, eliminating the need for customers to manage their own LLM provider keys. These developments signal that demand for GPU inference capacity in video workflows is accelerating, and facilities like SYD1 that can offer high-density Blackwell clusters in-region will attract streaming platforms seeking to run AI workloads without shipping data offshore.

In short

GreenSquare has partnered with Sharon AI to deploy 8,200 NVIDIA Blackwell Ultra GPUs at its SYD1 data center in Sydney. This project provides sovereign AI infrastructure for Australian customers starting in Q4 2026, establishing a new commercial benchmark for high-density AI hardware and brownfield data center redevelopment in the region.

FAQ

What hardware is being deployed at the SYD1 data center?

The SYD1 data center will house 8,200 NVIDIA Blackwell Ultra GPUs and HGX B300 systems.

When will the Sharon AI infrastructure be operational?

The infrastructure is scheduled to support model training and inference starting in Q4 2026.

Why is this deployment significant for Australian AI?

It provides a local sovereign alternative to global hyperscalers, allowing organizations to run AI workloads in-region without shipping data offshore.

How does the project address sustainability?

The project reuses an existing data center building, which helps reduce the carbon footprint and accelerates delivery timelines.


Read full article at telconews.com.au

Enjoy our coverage?

Add StreamingMeme as a preferred source on Google to see more of our streaming news at the top of your Search results.

Add as preferred source

Related Articles

Get this in your inbox → Subscribe