OPINION | AI Is Everywhere. But Why Is Nobody Talking About Infrastructure?

<p>Every week the world celebrates another breakthrough in Artificial Intelligence. A new large language model, a smarter AI assistant, a revolutionary image generator or an AI-powered product dominates the headlines. The discussion invariably revolves around software, what AI can write, analyse, diagnose or automate.</p>
<p>After 36 years of designing electronic products and building manufacturing ecosystems, I find myself asking a different question.</p>
<p>What does AI need before it can do any of those things?</p>
<p>The answer is surprisingly absent from most discussions. AI is not merely a software revolution. It is one of the largest infrastructure revolutions in modern industrial history.</p>
<p><iframe title="YouTube video player" src="https://www.youtube.com/embed/d1gX2Wemm4I?si=q0s4I67vJhhLnh6Z" width="560" height="315" frameborder="0" allowfullscreen="allowfullscreen"></iframe></p>
<h3><strong>The Biggest Misunderstanding</strong></h3>
<p>The term ‘AI infrastructure’ is often used as though it describes a single ecosystem. It does not.&nbsp;There are two entirely different infrastructures.&nbsp;The first creates Artificial Intelligence.&nbsp;The second deploys Artificial Intelligence.</p>
<p>The first is measured in gigawatts, semiconductor fabs and hyperscale computing.&nbsp;The second is measured in engineering innovation, embedded electronics, sensors and intelligent products.</p>
<p>Understanding this distinction changes how countries should invest, how industries should innovate and where companies like Unwired Connect fit into the AI economy.</p>
<p><strong>ALSO READ: <a href="https://news.abplive.com/business/india-top-family-businesses-hurun-2026-ambani-adani-1862487" target="_blank" rel="noopener">India&rsquo;s 300 Richest Family Businesses Worth Rs 138 Lakh Crore: Ambani Tops Hurun 2026 List</a></strong></p>
<h3><strong>Infrastructure For Generating AI</strong></h3>
<p>Training frontier AI models requires a pyramid of physical infrastructure.</p>
<p>At its base is reliable electricity. Modern AI data centres consume extraordinary amounts of power. Every watt entering a server eventually becomes heat, demanding sophisticated thermal management using liquid cooling, chillers, precision HVAC systems and water infrastructure.</p>
<p>Above power sits semiconductor manufacturing. GPUs, CPUs, NPUs, HBM memory, networking ASICs and power semiconductors depend on silicon wafers, lithography, deposition, etching, advanced packaging, testing and highly specialised materials.</p>
<p>Those devices are then assembled on complex multi-layer printed circuit boards through automated SMT production lines using SPI, pick-and-place systems, reflow soldering, AOI, X-ray inspection and functional testing.</p>
<p>Finally, thousands of servers are connected using optical networking capable of moving data between processors with almost no latency.</p>
<p>Only after this immense engineering stack exists can software engineers begin training AI models.</p>
<h3><strong>Where India Stands</strong></h3>
<p>India has become a global force in software engineering and semiconductor design, but leading-edge chip fabrication, advanced packaging, AI accelerators and hyperscale AI infrastructure remain concentrated in a handful of countries.</p>
<p>Acknowledging this is not pessimism. It is engineering realism.</p>
<p>Semiconductor ecosystems cannot be created overnight. They require decades of investment, policy continuity, academic collaboration, manufacturing capability and strategic global partnerships.</p>
<h3><strong>Infrastructure For Using AI</strong></h3>
<p>Fortunately, creating AI is only one side of the story. Deploying AI represents an equally important opportunity. Instead of trillion-parameter training clusters, deployment requires embedded processors, sensors, reliable connectivity, efficient power electronics and robust manufacturing.</p>
<p>Across India, AI is already transforming fintech, healthcare, logistics, agriculture, manufacturing and public infrastructure. Most of these systems rely on models created elsewhere but generate enormous value by solving local problems.</p>
<h3><strong>The Role Of Unwired Connect</strong></h3>
<p>This is where we have consciously positioned Unwired Connect.&nbsp;We are not building foundation models. We are embedding intelligence into infrastructure that already exists.</p>
<p>Lighting is one of the largest consumers of electricity in commercial buildings. Traditionally, a luminaire has been little more than a controlled electrical load. We believe it should become an intelligent decision-making system.</p>
<p>By combining IoT connectivity with embedded AI, lighting can understand occupancy, daylight availability and human activity. FMCW radar-based AI and vision-based AI enable luminaires to make decisions locally at the edge without relying on constant cloud connectivity or expensive computing infrastructure.</p>
<p>The outcome is practical rather than theoretical: lower energy consumption, improved occupant comfort and smarter buildings.</p>
<h3><strong>India’s Opportunity</strong></h3>
<p>India should pursue both infrastructures, but with different timelines.&nbsp;Long-term national investment should focus on semiconductor manufacturing, advanced packaging, AI-ready data centres, compound semiconductors, power electronics and world-class manufacturing capability.</p>
<p>Simultaneously, Indian companies should lead the world in AI-enabled products, Industrial IoT, edge AI, intelligent sensors, PCB manufacturing and automated electronics assembly. These are areas where India already possesses engineering talent and manufacturing experience capable of delivering global competitiveness.</p>
<p>Innovation does not happen only inside software companies. It also happens inside design laboratories, clean rooms, SMT lines, reliability laboratories and factories.&nbsp;History celebrates the applications that change society, but rarely acknowledges the invisible infrastructure that made those applications possible.</p>
<p>If India wants to become more than a consumer of Artificial Intelligence, it must patiently build the ecosystem required to generate AI while aggressively leading the world in deploying AI into products that improve everyday life.</p>
<p>The future of AI will not belong only to the countries that write the smartest algorithms. It will belong to those that master the complete engineering ecosystem, from electricity and semiconductors to manufacturing and intelligent products.</p>
<p>That is the conversation India should be having.</p>
<p><em>(The author is the Co-Founder and CTO of Unwired Connect)</em></p>
<p><em><strong>Disclaimer:</strong> The opinions, beliefs, and views expressed by the various authors and forum participants on this website are personal and do not reflect the opinions, beliefs, and views of ABP Network Pvt. Ltd.</em></p>



Source link