Skip to content
ISSUE 4 · HIGH-TECH MANUFACTURINGInfrastructure & Deep-Tech7 min read

Infrastructure Decided Intelligence.

The race for semiconductors and physical AI will be won in the concrete, not in the design lab.

Umesh Uttamchandani·Co-founder & Chief Growth Officer, DevX·2026
Umesh Uttamchandani, Co-founder & Chief Growth Officer, DevX

Umesh Uttamchandani

Co-founder & Chief Growth Officer, DevX

“India is funding outcomes at the end of the cycle. It should be funding attempts at the beginning—because without early risk, no new industrial capability gets built.”

— Umesh Uttamchandani

1M sq ft R&D facility cost: ₹400 crore. GST lost: ₹50 crore (12.5% capital hit).

American Express: shattered a 3,000-person campus into distributed nodes across Ahmedabad, Udaipur, Gandhinagar.

The constraint: not intelligence — the grid.

India’s deep-tech ambition has a physical limit. We project dominance in semiconductor design while ignoring the architecture required to host it. The sovereign R&D engine cannot run on venture capital alone; it requires uncompromising physical infrastructure.

The constraint is no longer intelligence. It is the grid.

Input Credit Distortion

Capital allocation currently rewards the finish line while starving the starting block. Constructing a standard one-million-square-foot R&D facility demands an injection of roughly ₹400 crore. Within that expenditure, nearly ₹50 crore—a direct 12.5% hit to capital efficiency—is lost to unrecoverable GST. By treating the architects of the industrial future identically to generic residential builders, the state cannibalises capital directly from the domestic supply chain.

“India is funding outcomes at the end of the cycle,” observes Umesh Uttamchandani, Co-founder of DevX. “It should be funding attempts at the beginning—because without early risk, no new industrial capability gets built.”

This policy loop favours incumbents while stalling the infrastructure required for the next generation of attempts.

Capital Cannot Enter Broken Systems

Global institutional liquidity is seeking deployment in India, but it encounters a “language mismatch.” Institutional consumers demand compliant, Grade-A assets, yet they are forced to negotiate with traditional landowners who lack the vocabulary of global standards.

“Tokenisation is the ultimate capital unlock,” Uttamchandani asserts.

By breaking down prime industrial assets into fractionalised ownership structures, global liquidity circumvents regulatory choke points. This shifts the ecosystem from a reliance on constrained captive capital to an influx of global institutional liquidity. Infrastructure is no longer a land-locked liability; it is a tradable security.

Talent Gravity: The Hub-and-Spoke Pivot

The narrative of Tier-2 cost arbitrage is dead. Organisations are decentralising because talent is no longer mobile. The modern engineer refuses the metropolitan commute; infrastructure must move to the talent. The American Express deployment—shattering a 3,000-person Gurgaon campus into distributed nodes in Ahmedabad, Udaipur, and Gandhinagar—is the definitive evidence of this shift.

However, decentralisation exposes the fragility of unorganised hardware ecosystems.

“Hardware does not fail gradually—it fails completely. One missed component in a sequencing chain can collapse the entire manufacturing cycle.”

Deep-tech cannot scale on broken logistics or Tier-2 power fluctuations. To anchor this talent, the planning sequence must be reversed: social infrastructure must be established before the real estate.

Compliance Barrier

As manufacturing integrates with global supply chains, European ESG standards have become a mandatory entry price. For mid-market hardware firms, this cost of compliance is skyrocketing. The grid must provide Integrated Compliance as a shared utility.

This requires embedding hardware solutions—smart HVAC monitoring and water metering—directly into the industrial shell to reduce and monitor “last-mile consumption.” Simultaneously, securing the physical grid is the prerequisite for data sovereignty. Adopting global technologies like Mistral AI is only viable if the data is hosted within the country.

“Technologies are being adopted from a global standpoint,” Uttamchandani notes, “but the data must be hosted locally to ensure sovereign intellectual property.”

India will not lose the deep-tech race on intelligence. It will lose it on infrastructure—unless the grid is built first.

Solutions Spotlight

  • Managed Industrial Units

    Stripping operational load by bundling real estate with integrated compliance, power redundancy, and talent recruitment.

  • Fractional Ownership

    Utilising tokenisation to allow global institutional capital to bypass fragmented land ownership.

  • Social-First Planning

    Reversing the urban sequence to establish social infrastructure as the lead domino for talent retention.

Sovereign Grid

  • The GST Tax

    Unrecoverable GST on industrial shells acts as a 12–15% tax on national manufacturing capability.

  • The Planning Reverse

    Social infrastructure must precede real estate to anchor high-value talent in regional hubs.

  • The Fragility Rule

    Logic fails in sequences. One missed component in a logistics chain collapses the entire cycle.

  • Data Locality

    Global tech adoption requires local data hosting to protect sovereign intellectual property.

Strategy Note

The Managed Factory Pivot

Managed workspaces are a legacy model. The new architecture is the managed factory. Securing deep-tech assets—semiconductors and physical AI—requires stripping the operational load from the hardware architect. Raw real estate is no longer the product; the grid must bundle integrated compliance, power redundancy, and talent logistics. India’s deep-tech victory will not be decided in the design phase. It will be won in the concrete.