Surprisingly Useful AI Article Enhancements
Lenovo has confirmed its involvement in delivering a 1,000-GPU Nvidia B200 cluster for Australian neocloud provider Sharon AI (NASDAQ:SHAZ), to be housed at NEXTDC’s Tier IV M3 data centre in Melbourne.
The deployment represents Lenovo’s largest TruScale Infrastructure as a Service engagement to date and forms part of a broader push to expand sovereign AI infrastructure across Australia.
Sovereign AI ambitions
The planned cluster is expected to give Australian and wider APAC enterprises, government organisations and research institutions access to accelerated computing resources without sending data offshore.
Sumir Bhatia, President of Lenovo Infrastructure Solutions Group, Asia Pacific, described the project as a step forward for sovereign AI capability in the region.
“Australia is well poised to be a global leader in AI, and Lenovo is proud to be enabling this next phase of sovereign AI capability,” Bhatia commented.
“By expanding our relationship with Sharon AI, we’re helping Australian organisations access AI compute on their own terms, securely and at scale.”
A milestone for local infrastructure
Sharon AI Co-Founder and CEO, James Manning, pointed to the significance of bringing a large-scale Nvidia Blackwell GPU cluster online in Melbourne.
“This deployment marks a major milestone for sovereign AI capability in Australia,” Manning explained. “In partnership with Lenovo, we are bringing one of the region’s most advanced Nvidia Blackwell GPU clusters online in Melbourne, giving Australian enterprises, government and researchers access to AI infrastructure on home soil.”
Manning went on to frame the project as being about more than raw compute power.
“Beyond scale, this is about national capability, data sovereignty and accelerating innovation locally, while positioning Australia as a serious contender in the global AI economy,” he added.
TruScale underpins flexible rollout
The deployment highlights how Lenovo’s TruScale Infrastructure as a Service model can support flexible and rapid infrastructure expansion as business needs shift.
By combining Lenovo’s high-density server engineering with VAST Data’s high-performance storage fabric, Sharon AI aims to offer a unified develop-to-deploy pipeline designed to reduce traditional bottlenecks in the AI lifecycle.
The as-a-service approach allows Sharon AI to scale its infrastructure without the capital expenditure typically associated with large GPU deployments.
Nvidia Blackwell architecture at the core
The cluster leverages Nvidia’s DGX Blackwell architecture, which is optimised for large-scale workloads including large language models (LLMs), recommender systems and real-time chatbots.
Each node is equipped with eight Blackwell GPUs interconnected via fifth-generation NVLink. According to Nvidia’s benchmarks, the architecture delivers up to three times the training performance and 15 times the inference performance when compared with previous-generation systems.
Related: Best Business Laptops for work & school
Related: Best Gaming Laptops
Related: Best Portable Laptop
Building on an existing footprint
The Melbourne project builds on Lenovo’s ongoing involvement in Australia’s AI and high-performance computing ecosystem.
It also expands Sharon AI’s existing local GPU fleet, which already includes Nvidia A40, L40s, H100 and H200 architectures.
Looking ahead, Lenovo has indicated it will continue to support Sharon AI’s infrastructure roadmap with next-generation B300 and GB300 platforms as they become available through 2026.
Last Updated on March 10, 2026 by Nick Ross
