Neysa deploys AMD Instinct MI350X GPU cluster in Mumbai for AI workloads

What happened
Neysa has launched a new GPU cluster in Mumbai using AMD Instinct MI350X chips, delivered through its Tytan AI infrastructure platform. The cluster is now operational and is designed to support training, fine-tuning, and inference for large language models and generative AI. Customers, both in India and internationally, have already contracted capacity on the cluster. The deployment gives users dedicated access to the hardware and control over how they configure and operate their AI workloads.
Key details
- The cluster uses AMD Instinct MI350X GPUs and AMD EPYC processors.
- Customers can choose deployment options including bare metal, Slurm, and Kubernetes.
- The exact number of GPUs and total compute capacity have not been disclosed.
- The cluster is part of Neysa's Tytan platform, which offers silicon-agnostic infrastructure.
- Neysa also operates Velocis, a managed AI cloud platform, and has announced plans for further GPU deployments in India.
Why it matters
Indian enterprises and global AI companies can now access dedicated, configurable GPU resources for advanced AI workloads in India, which can help meet data residency and latency requirements. Neysa customers gain more flexibility in how they deploy and manage AI infrastructure, with options for different hardware and software configurations. This deployment increases the available AI compute capacity in India, supporting both local and international demand for large-scale AI training and inference.

