Digital Realty and Vultr have entered into a strategic global partnership to deploy enterprise‑grade, GPU‑accelerated AI infrastructure in key data‑hub markets. Through this alliance, Vultr’s cloud‑based GPU clusters – powered by NVIDIA HGX B200 and AMD Instinct MI325X chips – will be integrated into Digital Realty’s PlatformDIGITAL data‑centre network, offering high‑performance AI compute close to where enterprise data resides.
The new offering allows organisations to access Vultr’s GPU resources via direct interconnection through Digital Realty’s ServiceFabric, using the AI Private Exchange for secure, low‑latency hybrid and multi‑party AI operations. Initial availability is live in Atlanta, Dallas, London and Singapore, with further roll‑outs underway in San Francisco, Frankfurt, Mumbai, Sydney and Tokyo.
The solution addresses three pressing enterprise needs: delivering immediate access to GPU clusters, ensuring compliance and data sovereignty through localised deployment, and offering flexible, consumption‑based pricing models without long‑term commitment burdens.
Chris Sharp, Chief Technology Officer of Digital Realty, emphasised that the partnership enables AI to operate “where enterprise data lives,” combining Vultr’s “price‑to‑performance model” with Digital Realty’s global infrastructure footprint. He highlighted that ServiceFabric and AIPx provide “secure orchestration pathways to distributed compute and multi‑party AI collaboration,” enabling enterprises to focus on outcomes rather than custom‑built environments.
Kevin Cochrane, Chief Marketing Officer of Vultr, said enterprises require AI infrastructure that is “powerful, but production‑ready responsive to compliance considerations.” He added that the collaboration delivers “immediate access to high‑performance GPU infrastructure in the world’s most strategic digital hubs,” enabling organisations to scale AI with confidence.
Vultr’s infrastructure depth complements Digital Realty’s expansive PlatformDIGITAL ecosystem, comprising more than 330 data centres across 55 metropolitan areas. The combined offering supports a variety of AI workloads, including model training, inference, agentic AI, analytics and virtual workstation deployment – all within pre‑validated, AI‑ready environments.
By aligning compute infrastructure close to data sources, enterprises can achieve lower latency, adhere to regulatory and sovereignty standards, and enjoy scalable performance without being locked into long‑term contracts. The partnership reflects growing enterprise demand for dedicated, compliant, and high‑density AI infrastructure that supports dynamic hybrid and multi‑cloud strategies.
Vultr, with its history of GPU expansion funded in part by AMD and LuminArx, has become a significant player in the “neocloud” ecosystem, distinguished from traditional hyperscalers by flexibility, proximity and cost‑efficiency. The company’s expanded GPU infrastructure and ambitious global footprint position it well to meet rising demand as enterprise use of AI evolves from training-focused workloads to inference and edge deployment.
Emerging enterprise architectures now increasingly favour composable AI stacks that can scale on demand. This partnership provides a ready-made infrastructure backbone combining colocation, proximity networking and GPU compute—a practical approach for businesses aiming to modernise legacy operations while maintaining control over data flow and cost.
Demand for AI infrastructure, particularly GPU-driven compute, continues to surge with enterprises across verticals—such as financial services, healthcare, media and e-commerce—seeking compliant, low-latency deployment models. The Digital Realty–Vultr alliance signals a market shift towards distributed, consumption-based AI infrastructure models that reduce complexity and support global hybrid operations.
Forthcoming expansions into Frankfurt, Mumbai, Sydney and Tokyo suggest both companies anticipate continued upward demand and regulatory complexity across jurisdictions. Their approach aligns with emerging trends towards sovereign compute zones and private AI platforms tailored for sensitive workloads.