Enterprise AI
Neysa has deployed an 8,192-GPU AMD Instinct MI350X cluster in Mumbai through its Tytan bare-metal AI infrastructure platform. The operational cluster is designed to support large language model training, fine-tuning, generative AI workloads and high-throughput inference. It combines AMD Instinct MI350X GPUs, AMD EPYC processors and the AMD ROCm software ecosystem.
According to Neysa, the cluster was brought online in a single phase, with global customers already having contracted capacity for AI deployments in India. Customers receive dedicated accelerator access and can configure network architecture, storage, deployment models and service-level requirements. Available options include bare metal, bare metal with Slurm, and bare metal with Kubernetes.
The deployment reflects Neysa's silicon-agnostic infrastructure strategy, allowing customers to select accelerator architectures and configurations while Neysa manages deployment and operations. The company also operates Velocis, its managed AI cloud platform. In October, Neysa partnered with STT GDC to deploy up to 9,000 GPUs at STT Noida 2, with the first phase expected in Q1 2027.
Why It Matters:
The deployment adds large-scale GPU infrastructure in India for organisations developing and running AI models. Dedicated accelerator access and configurable deployment options can support enterprises with different AI computing, storage and orchestration requirements.
CXO TechBOT Perspective:
As demand for AI computing grows, access to operational GPU capacity is becoming a critical consideration for enterprise AI strategies. Neysa's deployment highlights the role of domestic infrastructure providers in supporting large-scale AI development and inference workloads.