IREN Achieves NVIDIA Exemplar Cloud Status on NVIDIA HGX B300, Advancing Production-Grade AI Infrastructure

There is no shortage of performance claims in AI infrastructure. NVIDIA Exemplar Cloud is one of the few initiatives designed to verify them. IREN has achieved NVIDIA Exemplar Cloud on NVIDIA HGX B300 for training workloads. This status confirms that IREN's infrastructure performs within NVIDIA's reference performance targets across its full suite of benchmarking recipes, validated against NVIDIA reference architecture.

What Is NVIDIA Exemplar Cloud Status?

NVIDIA's Exemplar Cloud initiative evaluates cloud providers using its publicly available Performance Benchmarking recipes, which are end-to-end training workloads run across multi-node GPU clusters. The test suite covers large-scale pretraining runs on frontier model architectures, evaluating training throughput, networking efficiency, and cluster reliability under conditions that reflect real production workloads. To achieve this status, a provider must perform within NVIDIA's reference performance targets across every prescribed recipe for the hardware generation being validated.

A cluster that performs well in isolation does not always perform well at scale. Networking bottlenecks, suboptimal software stack configuration, and inconsistent cluster behavior under sustained load are the variables that can inhibit infrastructure delivery at production scale.

How Vertical Integration Supports Certified Production-Scale AI Training Performance

IREN is among a small group of cloud providers to achieve Exemplar Cloud status on NVIDIA HGX B300. This effort required close collaboration between IREN's engineering teams and NVIDIA performance experts across every layer of the stack, from the physical design of IREN's data centers through networking configuration, software tuning, and cluster operations.

Because the same organization that develops the sites, constructs the facilities, and deploys the hardware also operates the platform, there is a continuous feedback loop between physical infrastructure and workload performance. That feedback loop helps IREN to tune hardware and operations together, supporting stronger performance.

“Achieving NVIDIA Exemplar Cloud status is evidence of the performance and execution standards we are bringing to AI Cloud,” said Denis Skrinnikoff, Chief Technology Officer at IREN. “Customers running large training and inference workloads need infrastructure that performs predictably under load, and that is what we have engineered for. This milestone reinforces IREN’s position as a trusted AI Cloud provider as we continue scaling across our vertically integrated platform.”

"IREN's achievement of NVIDIA Exemplar Cloud status reflects deep engineering collaboration between our teams and the quality of infrastructure behind IREN’s AI Cloud, giving enterprises confidence to run their most demanding training workloads at scale," said Warren Barkley, VP Product Management, NVIDIA.

For teams building foundation models, fine-tuning domain-specific models, developing multimodal systems, running large-scale inference, accelerating scientific discovery, or executing simulation and analytics workloads, Exemplar Cloud status provides added confidence that the NVIDIA HGX B300 environment on IREN Cloud™ can support modern AI workloads. Impact to AI Builders and Enterprises with Long-Term AI Roadmaps For our customers, the goal is a more consistent and performant cloud experience.

By combining NVIDIA HGX B300 training capability with IREN’s vertically integrated data center platform, IREN strengthens its role as a trusted partner for organizations that need to scale AI with confidence. Run Training Workloads with NVIDIA HGX B300 on IREN Cloud™

For enterprises evaluating AI cloud infrastructure, Exemplar Cloud status provides an independently verified performance baseline on NVIDIA HGX B300. Explore IREN Cloud™ to learn more.