Compute AI Infrastructure for Abundant Intelligence
Curiosity AI is the full-stack compute infrastructure platform bringing megawatt-scale AI capacity online by partnering with AI Factories to serve large-scale neo-clouds and enterprise AI.
Infrastructure that's
already moving.
Partnered from day one
We work with AI Factories and AI Data Centres that already hold the power, land and institutional capability to scale.
Capital that deploys
Curiosity AI invests and co-invests directly in GPU and supporting compute infrastructure inside partner facilities.
Contracted from the start
Capacity is underwritten by long-term offtake: 3 to 5 year reservations from neo-clouds and enterprises.
The Stack for Scaling Compute
Five layers that turn powered land into contracted, revenue-generating AI compute.
Built on Proven Sites
Curiosity AI partners with AI Factories and AI Data Centres across India that already hold the power, land, capital and institutional capability to deploy 5MW to 100MW+ facilities in a phased manner, removing years of greenfield lead time.
Infrastructure as an Investment
The platform builds, invests and co-invests in GPU compute infrastructure and everything supporting it (CPUs, networking, advanced cooling, and storage) so partners never carry the full capital load alone.
Deployment as an Operating System
Curiosity AI leads and coordinates institutional-scale GPU deployment end to end, moving from a signed offtake agreement to a running, revenue-generating cluster without hand-offs between disconnected vendors.
Global Demand, Indian Capacity
The platform partners with large-scale neo-clouds to translate contracted U.S. demand into dedicated AI compute capacity located in India, aggregating demand centred on NVIDIA and AMD GPUs.
Sovereign Compute, Delivered
Scalable, reliable, sovereign GPU-as-a-Service and bare-metal compute is delivered to neo-cloud and enterprise customers for large-scale training workloads, inference and frontier AI.
Compute AI Infrastructure Platform for Abundant Intelligence.
Not another GPU cloud: the platform that brings megawatt-scale capacity online in the first place.
Everything in. Contracted compute out.
Land, power, capital, GPUs and supporting infrastructure flow into the platform. Dedicated compute capacity flows out to neo-clouds and enterprises.
The opportunity of
a generation.
Demand for AI compute is running far ahead of deployed supply. Curiosity AI exists to close that gap in India.
Compute is the constraint, not demand
Contracted backlog across leading global neo-clouds runs tens to hundreds of times ahead of current quarterly revenue. The bottleneck is deployed capacity.
AI infrastructure is an energy problem
Megawatt-scale compute needs power, land and connectivity before it needs GPUs. Curiosity AI is built around that sequence, not against it.
India is the next deployment frontier
AI-ready data centres in Mumbai and Chennai now have the power and institutional capability to host frontier-scale compute, along with the capital to scale it.
Long-term contracts underwrite the build
Investors value neo-clouds on multi-year infrastructure contracts. Curiosity AI converts that same contracted demand into physical capacity.
5MW to 100MW+, in phases.
First capacity online.
Air and liquid cooled fleet.
Additional partner sites.
National compute network.
From the buildout.
Phase 2 expansions: 10MW liquid-cooled NVIDIA GB300 roadmap
Four additional SuperPOD cluster of liquid-cooled NVIDIA GB300 GPUs join the air-cooled B300 fleet.
Phase 1 begins: 5MW of NVIDIA Blackwell B300 capacity
Four SuperPOD clusters enter deployment with a partner AI Factory on a five-year colocation model.
Scaling AI Compute Across Mumbai and Chennai
Curiosity AI partners with top Indian data centres to deliver institutional-scale GPU clusters.