Dell and NVIDIA combine to power NxtGen’s largest India AI factory

HIGHLIGHTS

Dell, NVIDIA tech powers NxtGen's largest India AI Factory

Blackwell GPUs, liquid cooling, and fast networks power AI training

This is the infrastructure behind India’s next AI wave

Dell and NVIDIA combine to power NxtGen’s largest India AI factory

For most people, the phrase “AI factory” sounds somewhere between a buzzword and a sci-fi prop. But in reality, it’s a tangible thing that India is starting to build at a serious scale.

Digit.in Survey
✅ Thank you for completing the survey!

That’s the context behind Dell Technologies and NVIDIA working with NxtGen to build what they’re calling India’s largest dedicated AI factory.

At its simplest, an AI factory is exactly what it sounds like – it’s essentially technology hardware infrastructure designed to mass-produce intelligence. Not apps, but trained AI models, again and again, at speed and massive scale. And that requires a very different kind of hardware stack than the data centres that power your regular email inbox or cloud storage.

At the heart of this setup is Dell’s AI Factory being built by NxtGen on top of the NVIDIA platform. In other words, Dell’s essentially offering a complete, pre-validated AI stack – full of compute, networking, storage, cooling – designed to work together without prolonged trial and error. NxtGen, which is in the business of building cutting edge data centres, is using Dell and NVIDIA hardware to build this AI factory to power AI workloads for predominantly Indian use cases.

The compute layer is doing the real heavy lifting. Dell is deploying PowerEdge XE9685L servers – machines purpose-built for AI training and inference – packed with NVIDIA’s latest Blackwell GPUs. We’re talking about thousands of GPUs working in parallel, designed to train large models faster, more efficiently, and at higher reliability than previous generations. Ultimately, this AI factory isn’t about running one chatbot, but about running many large workloads simultaneously without bottlenecks.

Also read: DGX Spark, NVLink Fusion, RTX PRO Servers – NVIDIA’s Full AI Stack Revealed

Of course, with a data centre labelled as AI Factory comes unprecedented cooling needs. GPUs at this scale generate enormous heat, and traditional air-cooled data centres simply don’t cut it. Cutting edge liquid-cooled systems deployed by NxtGen, allows the AI factory to push higher performance per rack while keeping power and thermal limits under control. This translates to more compute in less space, without cooking the hardware.

Networking is the other silent killer of AI performance, and it’s where NVIDIA’s Spectrum-X Ethernet platform comes into play. On top of that sits NVIDIA’s BlueField-3 DPUs, which offload networking, security, and data movement tasks away from the CPUs and GPUs. This frees up more compute for actual AI work – a small architectural detail that makes a big difference at scale.

Storage, meanwhile, is handled by Dell PowerScale systems designed to feed vast datasets into GPUs without becoming a choke point.  

What ties all of this together is Dell’s approach to integration. Instead of customers stitching together servers, networking, software, and cooling from multiple vendors, the AI factory model delivers a tightly integrated system that’s already tuned for AI workloads. For startups, enterprises, and institutions consuming this infrastructure as a service, that means faster deployment and fewer unpleasant surprises along the way.

The bigger takeaway here isn’t the headline GPU numbers – and for the record, it’s over 4000 NVIDIA Blackwell GPUs – it’s that India is finally building AI infrastructure that’s designed for AI from day one, not reimagined as an afterthought. This is the crucial plumbing behind the AI boom, and it’s where the real work happens.

You may never see this factory, nor will you be able to download it. But every serious AI system built on top of it will carry its fingerprints quietly in the background, unbeknownst to us all.

Also read: AI Factories to Agentic Web: NVIDIA and Microsoft’s vision for Future of AI

Jayesh Shinde

Jayesh Shinde

Executive Editor at Digit. Technology journalist since Jan 2008, with stints at Indiatimes.com and PCWorld.in. Enthusiastic dad, reluctant traveler, weekend gamer, LOTR nerd, pseudo bon vivant. View Full Profile

Digit.in
Logo
Digit.in
Logo