Why Is Nikhil Kamath Investing ₹200 Crore in Data Centres Like CtrlS in 2026?

When one of India's most closely watched investors puts ₹200 crore into a data centre company, it isn't a portfolio footnote — it's a signal. Nikhil Kamath's investment in CtrlS, paired with his line that 'every meaningful tech shift runs on' the infrastructure beneath it, points to a simple but often ignored truth: the AI boom everyone is chasing at the application layer is actually being won at the infrastructure layer.

What is the Concept

Data centres are the physical backbone of every digital product — the servers, storage, and networking that host websites, apps, AI models, and cloud platforms. CtrlS is one of India's established data centre operators, providing colocation and cloud infrastructure to enterprises that need reliable, high-density computing power. Kamath's ₹200 crore investment treats data centre capacity not as a utility cost but as a scarce, appreciating asset — similar to how early investors treated bandwidth during the dot-com buildout or cloud capacity during the 2010s SaaS wave.

The underlying idea is simple: software gets the attention, but infrastructure captures the value. Every AI model, every SaaS product, and every automation workflow ultimately runs on compute that has to physically exist somewhere. Investors betting on data centres are effectively betting on demand for AI and cloud services outpacing the supply of infrastructure that can serve it.

Why It Matters Now (2025–2026 Context)

India's AI adoption curve has moved faster than its infrastructure buildout. Enterprises are deploying more AI workloads, generating more data, and demanding lower latency — all of which require significantly more compute capacity than the traditional SaaS era. This mismatch between demand and supply is exactly the gap Kamath's investment is targeting, and it's the same gap founders and CTOs are quietly running into when they scale AI features and find cloud costs spiking unpredictably.

For founders, this matters because infrastructure scarcity translates directly into pricing power for data centre and cloud providers — which means rising compute costs for everyone building on top of them. Businesses that don't plan for this now will find themselves negotiating from a position of weakness later, when capacity is tighter and vendors have more leverage.

How AI Is Changing This

Traditional data centres were built for predictable, steady-state workloads — hosting websites, running databases, storing files. AI workloads are different: training and inference require dense clusters of high-performance compute that consume far more power and cooling per rack than legacy infrastructure. This is why investors like Kamath are backing data centre operators specifically, rather than generic cloud resellers — the physical capacity to run AI at scale is becoming the actual bottleneck, not the algorithms themselves.

This shift also changes who has negotiating power in the AI supply chain. A company that owns or has guaranteed access to AI-ready infrastructure can move faster and price more competitively than one that's simply renting whatever capacity is available on the open market. That's the contrarian insight most founders miss: the AI race isn't only about better models — it's increasingly about who secured compute capacity early.

Real-World Examples

Globally, this pattern has already played out. Microsoft's multi-billion-dollar commitments to data centre capacity to support OpenAI, and Amazon's continued expansion of AWS regions specifically to serve AI workloads, both reflect the same logic Kamath is applying at a smaller scale in India through CtrlS. In each case, the strategic move wasn't building a better model — it was locking in the infrastructure needed to run models at scale before competitors could.

Indian SMEs are already feeling a lighter version of this squeeze: companies running AI-powered customer support, recommendation engines, or document processing are seeing cloud bills grow faster than user growth, because compute-heavy workloads don't scale as cheaply as traditional web traffic did.

Practical Insights / Actions

Founders and CTOs should treat infrastructure planning as a strategic decision, not a DevOps afterthought. Use what we'd call the Compute Runway Framework: before scaling any AI feature, map out (1) current compute cost per active user, (2) projected cost at 10x scale, and (3) whether your provider can guarantee capacity at that scale without price spikes. Most teams only discover the answer to point three after they've already hit a wall.

The founder mistake to avoid here is optimizing for model performance while ignoring infrastructure cost curves — a fast, accurate AI feature that becomes unaffordable at scale is not a win. The hidden opportunity is the inverse: businesses that architect for compute efficiency early (batching, caching, right-sizing AI calls) can offer AI features profitably while competitors are still absorbing unpredictable cloud bills.

Future Outlook

Expect more capital to flow into Indian data centre infrastructure through 2026 and beyond, as AI adoption pushes compute demand higher across finance, retail, and healthcare. This will likely stabilize pricing for large enterprises that can commit to long-term capacity contracts, while smaller businesses relying on pay-as-you-go cloud pricing may continue to see cost volatility unless they actively manage their infrastructure strategy.

The broader signal from investments like Kamath's is that infrastructure is being re-rated from a cost center to a strategic asset — a mindset shift that founders building AI-driven products will need to adopt as well, or risk being priced out by competitors who planned ahead.

Conclusion

Kamath's ₹200 crore bet on CtrlS isn't really a story about one investor or one data centre company — it's a preview of where value is accumulating as AI adoption accelerates. Businesses that treat compute infrastructure as a strategic planning issue, rather than a line item to deal with later, will be better positioned to scale AI features profitably. If your team is scaling AI or automation workloads and unsure whether your current infrastructure setup can handle 2026-level demand, RP SoftTech can help audit your architecture and build a scalable, cost-efficient roadmap.