Vultr to deploy 20MW AMD MI355X cluster in Chennai, India

Vultr plans record 20MW AMD MI355X cluster in Chennai

By GPU Data Hub Desk · GPU Data Hub·Editorial standards·21 Oct 2026, 15:35·India

Vultr to deploy 20MW AMD MI355X cluster in Chennai, India

Cloud platform Vultr plans to install a 20MW cluster equipped with AMD MI355X accelerators in Chennai. According to the company, the installation will stand as the largest deployment of AMD Instinct hardware across India.

Original source

This summary was written by the GPU Data Hub desk from reporting published by Data Center Dynamics on 21 Oct 2026, 15:35. Read the full article at the original publisher.

Read Original Source →

Companies

Vultr AMD

Topics

AMD Instinct AI compute GPU deployment

LinkedIn X

More from GPU Data Hub

Latest →

Jensen Huang sends in the robot ducks to sell Nvidia's AI-everywhere vision

Nvidia chief executive Jensen Huang highlighted modular data centres, telecommunications infrastructure and robotics during a presentation in Berlin. The executive used the event to outline the chipmaker's strategy to deploy artificial intelligence technologies across diverse enterprise sectors. The keynote address was briefly disrupted by climate campaigners.

GPUs·3h ago·The Register

Nvidia’s latest GPU driver brings a new way to speed up PC games

Nvidia has launched driver version 621.01 to support Microsoft's Advanced Shader Delivery system for desktop gaming. The mechanism shifts shader compilation workloads away from client processors and into the cloud using DirectX 12. This approach is intended to accelerate initial title boot times and curb mid-game processing hitches across varied hardware configurations.

GPUs·4h ago·The Verge

Bringing the Spark to Europe: Local Intelligence Powers Next Wave of Agents

At GTC Berlin, Nvidia highlighted its comprehensive portfolio of on-premises hardware platforms configured for agentic AI workloads. The offering includes RTX Spark for Windows desktops, DGX Spark for Linux software engineers, and DGX Station platforms aimed at high-end enterprise tasks.

GPUs·7h ago·NVIDIA Newsroom

Nvidia GTC Berlin keynote disrupted by Greenpeace activists

Greenpeace campaigners staged a demonstration during the GTC Berlin keynote speech delivered by Nvidia chief executive Jensen Huang. The activists interrupted the presentation to protest against the semiconductor firm's environmental footprint, calling on the company to reform its operations.

GPUs·7h ago·Data Center Dynamics

NVIDIA cuDSS Now Generally Available to Accelerate Engineering and Scientific Simulations

NVIDIA has released cuDSS into general availability, introducing a CUDA-X library designed to execute direct sparse linear solvers directly on graphics hardware. The software aims to accelerate demanding computational simulations across sectors including semiconductor development, aerospace engineering, and industrial machinery.

GPUs·7h ago·NVIDIA Newsroom

Tens of thousands of Nvidia GPUs sit idle in Meta's Prometheus supercluster, just waiting for a failure

Meta has kept tens of thousands of Nvidia graphics processors on standby within its Prometheus supercluster to handle routine hardware breakdowns. The reserve inventory forms part of the company's strategy to maintain continuous uptime across an estate targeting one million active GPUs during training workloads.

GPUs·21h ago·The Register