Q1. Could you start by giving us a brief overview of your professional background, particularly focusing on your expertise in the industry?

With more than 26 years in engineering, project management, and leadership, my career has centered on mission-critical infrastructure—especially in Semiconductor Manufacturing Facilities (FABs and OSAT) and hyperscale Data Centers.

I have hands-on experience delivering large-scale greenfield and brownfield projects, guiding them through every stage—from early concepts and detailed design, all the way to procurement, construction, and commissioning—within both EPC/EPCM and PMC frameworks.

Throughout my career, I’ve had the opportunity to work with industry leaders like Larsen & Toubro, Voltas (a Tata Enterprise), Godrej, Jacobs, and ISYX Technologist. Currently, I lead Data Center & Semiconductor Projects at L&T. In these roles, I’ve guided diverse teams and overseen complex, high-value projects, always aiming to meet the highest standards of reliability, sustainability, and performance.

On the technical front, I specialize in MEP systems, cleanroom environments, and high-purity utilities for mission-critical settings. My work is further strengthened by expertise in techno-commercial evaluation, BIM-based digital delivery, and effective stakeholder management.

Academically, I hold a Ph.D. in Semiconductor Engineering, along with advanced degrees in Electrical Engineering and Operations Management. I’m also PMP-certified, a certified Energy Auditor, CDCP, and a Green Building Specialist.

Ultimately, what sets me apart is my ability to blend deep technical knowledge with practical, strategic leadership—delivering high-quality, cost-effective, and future-ready infrastructure for the semiconductor and data center industries.

 

Q2. From an execution standpoint, how heavily does the capital expenditure of constructing dedicated on-site power generation erode the ultimate IRR (Internal Rate of Return) of a 100 MW+ hyper-scaler asset?

From an execution standpoint, dedicated on-site power generation for a 100 MW+ hyperscale facility can account for roughly 20–30% of total CAPEX, and can initially compress IRR by about 150–250 basis points.

However, in mission-critical infrastructure, this should be viewed as a strategic investment rather than pure cost. On-site power ensures uptime reliability, directly protecting revenue and avoiding SLA penalties. It also enables faster time-to-market by reducing dependency on grid readiness—often offsetting IRR impact through earlier cash flows. Additionally, it enhances tenant confidence and allows premium pricing, particularly for hyperscalers with stringent availability requirements. 

My view is that, while it may dilute short-term returns, when executed with a phased, hybrid energy approach, it strengthens long-term asset value and stabilizes IRR.

 

Q3. Advanced AI data centers and semiconductor Fabs require millions of gallons of water daily. In hyper-arid or water-stressed regions like the Middle East or parts of India, what specific alternative design configurations are mandatory?

In regions like the Middle East and parts of India where water is scarce, traditional water-heavy designs simply aren’t practical anymore. For both AI data centers and semiconductor fabs, alternative approaches have become essential.

Key measures include:

• Using air-cooled or hybrid cooling systems to drastically cut down on the need for cooling water
• Implementing Zero Liquid Discharge (ZLD) systems in fabs, which can recover up to 90–95% of water used
• Maximizing recycling and cascading reuse throughout process and utility systems
• Switching to treated wastewater (TSE/STP) instead of relying on potable water
• Adopting dry or optimized process technologies to keep water usage in fabs to a minimum
• Setting up on-site desalination plants in coastal areas, particularly in the Middle East
• Leveraging digital water management systems for real-time monitoring and better efficiency

Water needs to be treated as a core design factor. Facilities should be built around high-efficiency, nearly water-neutral principles, making long-term sustainability and day-to-day operations possible even in the most water-challenged environments.

 

Q4. What are the primary technical barriers that make retrofitting a facility for high-density AI clusters economically unfeasible, forcing a pivot to a fresh Greenfield build?

Upgrading older facilities to support high-density AI clusters is often not practical because of basic limitations in power, cooling, and building structure.
Power density gap: Most legacy designs only handle about 5–10 kW per rack, but AI applications need 50–100 kW or more. Meeting these demands would require major electrical upgrades.

Cooling limitations: Traditional air-based cooling systems just can’t keep up with these higher densities, and switching to liquid cooling would mean a major redesign.

Structural constraints: Higher rack and equipment load exceed the original floor capacity

Space and layout issues: Limited room for additional power and cooling infrastructure

Operational disruption: Retrofitting in live environments increases risk, cost, and timelines

Economically, retrofit costs can reach 70–80% of a new build's cost without delivering the same scalability or efficiency.

Beyond a certain threshold, Greenfield development is the more viable option, allowing purpose-built design for AI workloads with better long-term performance and ROI.

 

Q5. High-voltage transformers, switchgear, and specialized liquid chillers are experiencing highly volatile lead times. From an active procurement standpoint, how are these component shortages stretching project timelines?

Shortages of key components—especially high-voltage transformers, GIS/switchgear, and large-capacity chillers—are causing major delays in project schedules, both when it comes to procurement and actually getting work done on site.

Key impacts include:

Longer wait times for critical equipment

For example, transformers and switchgear that used to take 6–9 months to arrive are now taking 12–18 months or even longer, while specialized chillers can take 40–60 weeks. These delays directly impact critical project steps.

Front-loaded procurement requirement

Orders have to be locked in much earlier—sometimes as soon as the design is finalized or even while it’s still being developed—which makes it harder to adjust plans later on.

Phasing and sequencing disruptions

If these long-lead items are delayed, it holds up energization, commissioning, and even partial handovers—pushing out revenue schedules.

Cost escalation and vendor risk

Prices are jumping around and there aren’t enough suppliers, so teams often have to re-bid, redesign, or change specifications on the fly.

Q6. From a pure infrastructure and facilities design standpoint, what are the incremental complexities when constructing a facility for advanced 2.5D/3D packaging compared to a traditional, legacy back-end OSAT assembly line?

Designing infrastructure for advanced 2.5D and 3D packaging is much more complex than working with traditional OSAT assembly lines. The higher level of precision, sensitivity, and utility needs bring a whole new set of challenges.

Key incremental challenges include:

Ultra-tight environmental control: The requirements for temperature, humidity, vibration, and AMC (Airborne Molecular Contamination) are much stricter, all to protect the delicate fine-pitch interconnects and TSV structures.

Advanced cleanroom standards: Facilities need higher cleanroom classifications, with localized mini-environments and even tighter particle control than before.

High-precision process tools & utilities: These processes need ultra-stable UPW, specialty gases, and chemical delivery systems, all with tighter tolerances and more built-in redundancy.

Thermal and process cooling demands: There’s a much greater need for high-accuracy process cooling systems, especially to support bonding, lithography, and advanced packaging operations.

Vibration and structural sensitivity: Tools like hybrid bonders are extra sensitive, so they demand better vibration isolation and stricter building design standards.

Integration of diverse process technologies: Facilities now have to combine front-end processes like lithography and etching with back-end assembly, which makes everything more complex.

Higher capex and tighter tolerances in design execution: All of this means projects require more coordination, precision engineering, and BIM-driven integration—raising both costs and the bar for execution.

 

Q7. If you were an investor looking at companies within the space, what critical question would you pose to their senior management?

How are you structurally positioning your infrastructure to remain scalable and future-ready, while protecting returns against rapid technology evolution and utility constraints?

I would ask this question because in this sector, success depends on the ability to balance high upfront CAPEX with long-term adaptability. I would assess how they approach modular scalability, phased investments, resilience of power and water infrastructure, supply chain security for critical equipment, and alignment with evolving ESG and regulatory requirements. Ultimately, the differentiator is not just building capacity, but building infrastructure that can adapt to changing technology and tenant demands while ensuring stable, predictable returns over the asset lifecycle.