VULTR launches 20 MW AI supercompute cluster in Chennai. Key Highlights from the launch and Leadership Remarks வரலாறு அறிவீர் 1. Vultr's AI Supercompute Cluster in Chennai Vultr has expanded its AI infrastructure footprint in India with the launch of a 20-megawatt AI supercompute cluster in Chennai, powered by AMD Instinct MI355X GPUs. Key details of the facility include: 20 MW computing capacity, with more than 10,000 AMD Instinct MI355X GPUs. | Powered by AMD Instinct MI355X GPUs and AMD EPYC CPUs to support large-scale AI training and inference workloads. | The Chennai cluster is Vultr's second-largest supercompute cluster globally, after its largest facility in the United States. | The company plans to continue expanding its computing infrastructure in India, bringing successive generations of GPUs and CPUs closer to developers and businesses. 2. Vultr's India Presence Vultr began its India operations in 2022, establishing three data centre regions in New Delhi NCR, Bengaluru and Mumbai. Its infrastructure initially leveraged AMD EPYC processors, with the company subsequently expanding its offering to AMD Instinct GPUs, including the MI300X, MI325X and now MI355X. Vultr is the world's largest privately held independent cloud infrastructure provider, with a global footprint spanning six continents. Its globally distributed network is designed to connect up to 90% of the world's population within two milliseconds, enabling developers and enterprises to access high-performance cloud infrastructure closer to where they operate and innovate. 3. StartupTN-Vultr MoU: Enabling Tamil Nadu's Startup Ecosystem Vultr India signed a Memorandum of Understanding (MoU) with StartupTN, the Tamil Nadu Startup and Innovation Mission, to extend cloud computing credits and preferential pricing to eligible startups across the state. The initiative aims to reduce infrastructure costs and improve access to computing resources for startups developing and scaling technology products. Benefits announced under the initiative: Early-stage startups: US$2,500 worth of Vultr credits, along with a 20% discount. | Eligible startups beyond the early-stage funding threshold: US$45,000 worth of Vultr credits, along with a 20% discount. The initiative is intended to support product development, experimentation and deployment, particularly for startups working in AI and deep technology. 4. Key Leadership Remarks Kevin Cochrane, Chief Marketing Officer, Vultr Kevin Cochrane highlighted India's growing importance as a global centre for AI development and the country's engineering talent as a key driver of Vultr's investment in the market. He noted that India is moving beyond consuming AI technologies to actively building and developing AI applications for global markets. He also highlighted Vultr's efforts to strengthen its developer relations function and support India's open-source developer ecosystem. On the Chennai investment, he emphasised the importance of bringing computing infrastructure closer to the communities driving innovation. "India isn't just consuming AI. The buildout of AI is happening here." "This is where the talent is. This is where the challenge is literally changing the world." Thiru. P Mathan Raja, Minister of Micro, Small & Medium Enterprises, Government of Tamil Nadu The Minister highlighted the importance of making advanced computing infrastructure accessible to startups and MSMEs across Tamil Nadu, including those in smaller cities. He noted that access to computing resources could support applications such as product design, software testing, inventory management and operational efficiency. He welcomed the MoU between Vultr and StartupTN, emphasising the need for affordable technology access to ensure that infrastructure investments translate into opportunities for businesses across the state. "The true measure of technology is not how sophisticated it is, but the opportunity it creates." Thiru. Dr. Atul Anand, IAS, Additional Chief Secretary to Government of Tamil Nadu, IT & Digital Services Department Dr. Atul Anand welcomed Vultr's investment in Chennai, highlighting Tamil Nadu's engineering talent, growing startup ecosystem of approximately 16,000 startups and established technology capabilities. He emphasised that affordable computing infrastructure is essential to accelerating AI innovation and encouraged Vultr to extend accessible pricing packages to startups, particularly those based in Tamil Nadu. He also assured the company of the state government's support for its operations and growth. வரலாறு அறிவீர் Thiru Singai G. Ramachandran Chairman, Tamil Nadu Startup & Innovation Mission Thiru Singai G. Ramachandran highlighted Tamil Nadu's growing manufacturing, electronics and deep-tech ecosystem, noting that access to advanced computing infrastructure is essential to taking homegrown innovation to global markets. The address highlighted Chennai's approximately 567 deep-tech startups, with an estimated ecosystem valuation of ₹75,000 crore, as well as the need to strengthen AI and software capabilities alongside the state's established manufacturing strengths. Emphasising the importance of measuring the impact of infrastructure investments, the StartupTN leadership noted: "Beyond measuring infrastructure by megawatts, measuring its impact is very important." Thiru Singai G. Ramachandran also highlighted the MoU with Vultr, including the cloud credits and preferential pricing being extended to eligible startups.

Gartner IT Symposium: key takeaways on the future of enterprise AI. Enterprise AI is moving beyond experimentation, but turning early progress into production-scale business impact requires more than access to GPUs. That was a central theme at Gartner IT Symposium, where Vultr joined conversations with industry analysts and enterprise technology leaders about the evolving AI infrastructure market, the growing role of CPUs, and what organizations need to deploy AI with trust, control, and measurable results. Across these discussions, one message stood out: Enterprises need infrastructure built around the applications they want to run, not just the accelerators they want to buy. Here are three key takeaways from this year's Symposium. 1. Enterprise AI needs more than GPU capacity. The rise of neocloud providers and the growing investment in AI infrastructure were major topics at Symposium. As organizations commit more resources to AI, questions about infrastructure economics, long-term value, and the ability to translate investment into business outcomes are becoming increasingly important. Gartner's October 2026 announcement that more than $3.5 trillion in AI infrastructure spending could be impaired by the end of 2030 added urgency to these conversations. The announcement underscored the need for organizations to evaluate AI infrastructure decisions carefully as the market evolves. For Vultr, the opportunity extends beyond providing GPU as a service (GPUaaS). Enterprises need a complete cloud infrastructure foundation that supports the full lifecycle of AI applications, including compute, accelerated infrastructure, storage, networking, and security. During Symposium, Vultr emphasized the importance of bringing together the strengths of traditional cloud computing and accelerated computing to help organizations move from individual AI experiments to real applications and production workloads. That broader approach also means giving enterprises flexibility to choose infrastructure based on their technical and business requirements, rather than assuming every AI workload needs the same architecture. 2. CPUs remain critical as agentic AI evolves. While GPUs continue to attract significant attention, the role of CPUs was another important theme throughout Symposium. In his discussions about the changing demands of AI infrastructure, Vultr CMO Kevin Cochrane reinforced a simple point: The CPU never went away. As enterprises deploy AI agents, the work extends beyond model training and inference. Agents need to coordinate tasks, call tools, retrieve information, interact with existing systems, and execute multi-step workflows. These activities place additional demands on CPU resources alongside GPUs. As Cochrane discussed in a recent interview with Fierce Network, Vultr is seeing demand for CPU resources increase as agentic AI workloads become more prevalent. The infrastructure requirements for enterprise applications differ from those of environments focused primarily on supplying large quantities of GPU capacity. For enterprises, the objective is to run complete applications that solve real problems. That requires balancing accelerated computing with the CPU performance and supporting infrastructure needed to coordinate and execute workloads. Vultr's AMD EPYC-powered VX1 Cloud Compute is part of this broader approach, providing dedicated CPU resources for the application logic and orchestration that support modern AI workloads. The takeaway from Symposium was clear: As AI moves into production, organizations must account for the entire workload rather than focusing exclusively on GPUs. 3. Start with the workload and measure the impact. Vultr's Piyush Shukla joined a customer panel moderated by HPE, alongside Sudheer Matta, SVP of HPE Networking, and Laura Lincoln of Grant Thornton. The discussion explored how organizations can scale AI with trust, control, and measurable impact. One of the central takeaways was to start with the workload and work backward. Before choosing a model or investing in infrastructure, organizations need to understand the business problem they are trying to solve. That means identifying the desired outcome, determining the application's requirements, and selecting the resources and architecture best suited to deliver the result. This approach helps enterprises make more informed infrastructure decisions. Some workloads may require substantial GPU capacity, while others depend more heavily on CPU performance, data access, storage, or networking. Many production applications require a combination of these capabilities. The panel also highlighted the importance of metrics. Deploying an AI application is only the beginning. Organizations need to establish how they will measure success, demonstrate business impact, and determine whether an initiative is delivering the expected results. Whether the objective is to improve operational efficiency, accelerate a business process, or enable a new capability, measurable outcomes provide a foundation for making AI investments sustainable over time. Trust and control are equally important. As AI agents gain the ability to interact with enterprise systems and take action, security must be part of the infrastructure conversation from the beginning. Organizations need the appropriate controls to protect sensitive workloads while enabling teams to move forward with AI adoption. From Symposium conversations to enterprise AI infrastructure. These discussions reflect the direction of Vultr's ongoing investments in enterprise AI. At Symposium, Vultr's broader enterprise AI developments and its Chennai expansion provided additional context on how the company is building infrastructure to support organizations as they move AI into production. The Chennai cloud data center region expands Vultr's global footprint and access to infrastructure for AI and enterprise workloads. Meanwhile, Vultr's enterprise AI platform developments focus on giving organizations the flexibility to combine compute, accelerated infrastructure, and supporting technologies according to their needs. Together, these efforts reflect a broader view of what enterprises require from their infrastructure providers: global access, workload flexibility, performance, and the security and control necessary to operate AI applications in real-world environments. As enterprise AI adoption advances, the buying conversation is expanding as well. Infrastructure decisions increasingly involve security leaders alongside IT, engineering, and business stakeholders. That makes a comprehensive approach to architecture and governance essential for organizations preparing to scale AI. The next phase of enterprise AI. Kevin Cochrane used an analogy from his experience running ultramarathons to describe the current stage of AI adoption: Enterprises have reached the end of the first marathon, but the journey is far from over. Organizations are making meaningful progress, moving beyond experimentation and beginning to put AI to work. The next challenge is building the infrastructure and operational foundations needed to scale those initiatives and deliver lasting business value. The conversations at Gartner IT Symposium reinforced several priorities for that next phase: Look beyond GPU capacity, account for the growing role of CPUs, start infrastructure decisions with the workload, and measure success by business outcomes. For Vultr, these priorities inform an approach to cloud infrastructure that combines accelerated computing with the broader capabilities enterprises need to build, deploy, and scale AI applications. The first phase of enterprise AI has demonstrated what is possible. The next will be defined by how effectively organizations implement it.

Vultr to Deploy a 20MW AMD AI Supercluster in Chennai, India. October 26, 2026 Vultr CEO J.J. Kardwell announces a 20MW data centre expansion in India using AMD GPUs to meet rising data sovereignty and local AI inference demands Vultr recently announced that it is expanding its data centre footprint in India by building a 20MW supercluster in Chennai using AMD Instinct MI355X GPUs. Each GPU features 288GB of HBM3E and 8TB/s of memory bandwidth to handle complex AI model development and deployment. The facility has been designed to meet the needs of AI labs and foundation model builders, AI-native startups, enterprises, and research and academic institutions. According to Vult, its Chennai deployment will join the company's supercluster in Springfield, Ohio, as one of the two largest AMD AI infrastructure projects globally. It will support sovereign cloud initiatives and demanding AI workloads, such as LLM and foundation model training, fine-tuning and high-throughput inference. Meeting local data residency rules. "Enterprises deploying AI at scale need infrastructure they can access locally to meet data sovereignty, data residency and latency requirements," says J.J. Kardwell, CEO of Vultr. "Vultr's Chennai supercluster, powered by AMD MI355X GPUs, gives organisations in India the capacity to train, fine-tune and run inference on in-country infrastructure." Enterprises deploying AI at scale need infrastructure they can access locally to meet data sovereignty, data residency and latency requirements J.J. Kardwell, CEO of Vultr "Enterprises deploying AI at scale need infrastructure they can access locally to meet data sovereignty, data residency and latency requirements," says J.J. Kardwell, CEO of Vultr. "Vultr's Chennai supercluster, powered by AMD MI355X GPUs, gives organisations in India the capacity to train, fine-tune and run inference on in-country infrastructure." Open software ecosystem access. India has the talent and ideas to build the next generation of AI Jaya Jagadish, Country Head & SVP Engineering at AMD India The Chennai supercluster uses the AMD ROCm open software ecosystem to give developers access to widely used AI frameworks, libraries and models. "India has the talent and ideas to build the next generation of AI," says Jaya Jagadish, Country Head & SVP Engineering at AMD India. "Access to compute at this scale will help turn more of those ideas into reality. "Through our collaboration with Vultr, we're expanding access to AMD Instinct compute in India, backed by the AMD ROCm open software ecosystem, giving researchers, startups and enterprises greater choice and flexibility in how they develop and deploy AI." Expanding rack-scale server architecture. The Chennai deployment is an extension of Vultr's relationship with AMD, which spans the AMD Instinct GPU roadmap, Vultr intends to significantly increase its capacity in 2027 using AMD Instinct MI455X GPUs and AMD Helios rack-scale AI architecture. Key deployment facts and partners. * Vultr is constructing a 20MW AI supercluster in Chennai. * The new infrastructure features AMD Instinct MI355X GPUs. * Each installed GPU delivers 288GB of HBM3E memory and 8TB/s of memory bandwidth. * AMD: Vultr collaborates with the hardware manufacturer to supply Instinct GPUs and EPYC data centre CPUs for its global enterprise cloud regions. * NVIDIA: Vultr partners with the hardware provider to offer widespread access to cloud-based enterprise GPUs. * PCCW Global's Console Connect: Vultr works with the networking platform to provide high-performance direct connections to its cloud infrastructure. Company Portals