SCX.ai Partners With DDN on Sovereign AI Cloud
SCX.ai and DDN announced a partnership on August 24, 2026, to expand an Australian sovereign AI inferencing cloud aimed at enterprises, government agencies, and research institutions. The companies are combining SCX.ai's domestic SambaNova-based infrastructure with DDN's Infinia data platform. CFOtech reports that the announcement followed SCX.ai's AUD $40 million ASX initial public offering and that its first sovereign AI node operates in Sydney.
SCX.ai and DDN announced a partnership on August 24, 2026, to expand what the companies describe as Australia's largest sovereign AI inferencing cloud. The integration combines SCX.ai's Australian infrastructure with DDN's Infinia data platform for enterprises, government agencies, and research institutions seeking to keep AI workloads and data onshore.
According to Unite.AI, the announcement came three days after SCX.ai began trading on the Australian Securities Exchange following a fully underwritten AUD $40 million IPO. CFOtech reports that SCX.ai had AUD $6.5 million in contracted annual recurring revenue at the end of July, up 20.9% from May, and more than 400 active users.
Data layer for inference workloads
The partners are targeting multi-tenant inference use cases, including workloads with large context windows and agentic AI patterns. According to the companies, Infinia provides sub-millisecond latency and can load key-value cache data up to 27 times faster. KV cache loading can become an I/O constraint in autoregressive inference when systems repeatedly access prior-token state, particularly under high concurrency or long-context workloads.
SCX.ai founder and CEO David Keane told Unite.AI: "For the first time, Australian organisations have access to a fully domestic, enterprise-grade AI cloud that scales effortlessly." He added that integrating DDN's platforms was intended to support data-intensive workloads.
CFOtech reports that SCX.ai's first sovereign AI node is operating at the Equinix SY5 data centre in Sydney. The reported domestic deployment focus reflects demand from organisations that want AI workloads and data to remain in Australia.
ASIC infrastructure and efficiency claims
Both reports state that SCX.ai uses SambaNova SN40L AI processors rather than GPUs and relies on air cooling. Ahead of its ASX listing, SCX.ai disclosed testing that it said showed roughly 2.5 to 5.6 times the performance per watt of GPU-based systems across selected stable inference workloads, depending on the tested model and hardware configuration.
Those figures are company-disclosed test results, so practitioners evaluating the infrastructure would need workload-specific comparisons covering model support, throughput, latency, cost, and operational availability. More broadly, AI infrastructure providers pursuing on-premises or sovereign deployments commonly emphasize power density, cooling constraints, and data locality because these factors can shape both capacity planning and regulatory fit.
The partnership adds a data-management layer to SCX.ai's inference infrastructure as Australian demand grows for locally hosted AI capacity.
Key Points
- 1SCX.ai and DDN combine domestic inference infrastructure with Infinia, targeting Australian organisations that require onshore AI workloads and data.
- 2The companies cite sub-millisecond latency and 27-times faster KV cache loading, metrics relevant to long-context and concurrent inference evaluation.
- 3Sovereign AI deployments commonly make data locality, power efficiency, cooling design, and workload-specific performance validation central procurement considerations.
Scoring Rationale
The partnership is a notable Australian AI infrastructure development, combining a sovereign inference operator with a data platform for enterprise and public-sector workloads. Its practitioner relevance is strongest for teams evaluating data-residency-constrained inference, though the reported performance figures remain company-disclosed rather than independently benchmarked.
Sources
Primary source and supporting 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
