Neysa deployed 8,192 AMD GPUs in Mumbai to boost AI infrastructure, the company said on November 18. This setup uses AMD Instinct MI350X accelerators to support large language models and generative AI. "Global customers have already contracted this capacity to support large-scale AI deployments in India," Neysa said about the project.

Mumbai-based AI cloud service provider Neysa has deployed an 8,192-GPU cluster powered by AMD's Instinct MI350X accelerators in Mumbai, expanding its dedicated AI computing infrastructure amid growing demand for large-scale AI workloads in India. Global customers have already contracted the capacity to support their AI deployments in India, the company said in an announcement on November 18. The deployment is powered by AMD Instinct MI350X GPUs, AMD EPYC processors and the AMD ROCm open software ecosystem. It is designed to support a range of workloads, including large language model (LLM) training, fine-tuning, generative AI applications and high-throughput inference.

The move also expands the range of accelerator options available on Neysa's Tytan platform as the company pursues a silicon-agnostic infrastructure strategy, allowing customers to choose computing hardware and configurations based on their requirements. Under this model, customers can determine their preferred accelerator architecture, network design, storage platform and deployment model, while Neysa manages cluster deployment and operations. "Global customers have already contracted this capacity to support large-scale AI deployments in India," Neysa said, adding that demand is increasingly shifting beyond GPU availability to infrastructure reliability, operational control and certainty of computing capacity.

How Neysa plans to scale its AI infrastructure Neysa Tytan is designed to provide dedicated bare-metal infrastructure for organisations running AI workloads at scale. Customers can deploy the infrastructure in different configurations, including bare metal, bare metal with Slurm, or bare metal with Kubernetes. The platform is aimed at organisations seeking dedicated access to computing hardware for training, fine-tuning and large-scale AI inference, rather than shared cloud infrastructure. "The question for large AI deployments is no longer simply how many GPUs are available. It is whether the infrastructure can be deployed at scale, operated with certainty and configured for the workload," said Sharad Sanghi, co-founder and chief executive officer of Neysa.

With the addition of AMD's MI350X GPUs, Neysa is looking to expand the choice of silicon available to customers while retaining a single operating layer for infrastructure management. Vinay Sinha, India managing director, sales at AMD, said the collaboration would expand access to advanced AI infrastructure in India. "Organizations are looking for flexibility in how they build and scale AI, and open ecosystems play an important role in enabling that choice," Sinha said. He added that the MI350X GPUs are designed to support a broad range of AI workloads, allowing customers to align infrastructure decisions with their performance and operational requirements.

The deployment comes amid growing efforts by technology companies and cloud providers to expand AI computing capacity in India.