India Semiconductor Strategy Must Go Beyond Chip Packaging, Says Rajeev Chandrasekhar

India has spent the last few years celebrating every new chip packaging plant as a victory — but the India semiconductor strategy debate took a sharp turn on October 2, when the man who helped build the country’s semiconductor mission basically said: that’s not enough.

Rajeev Chandrasekhar — former Minister of State for Electronics & IT and part of the team behind the India Semiconductor Mission, the DLI scheme and IndiaAI — has argued that the India semiconductor strategy needs to move beyond packaging and start building genuine technological capability. His prescription is specific, and it revolves around two words: AI inference.

India Semiconductor Strategy: Packaging Is the Warm-Up, Not the Game

Chandrasekhar’s critique is sharp. “The path of least resistance in industrial policy is to approve what can be approved quickly and measure success by investment commitments rather than by the strategic position those investments create,” he wrote. Translation: counting approved factories is easy; building the ability to design world-class chips is hard — and it’s the hard part that matters.

His framing is historical. “For decades, India has missed or been a distant player in major global industrial supercycles. This supercycle, we do not intend to miss,” he said, referring to the AI age. Packaging plants — where finished chips are assembled and tested — are the low-hanging fruit. Design IP, process technology and talent are the actual prize.

The Big Bet: AI Inference Chips

Here’s where it gets interesting. Chandrasekhar identifies AI inference — the process of running trained AI models, as opposed to training them — as India’s opening. His logic: AI training is increasingly locked up around Nvidia’s CUDA ecosystem and custom chips built by hyperscalers. That’s a fortress. But inference? “Inference is wide open,” he wrote.

He’s got a point. Inference workloads are wildly diverse — cloud servers, edge devices, defence systems, agricultural sensors, industrial equipment. No single architecture dominates, which means there’s room for specialised new chips. And India has raw material: 1.25 lakh chip-design engineers, one of the largest pools in the world.

His proposed weapon is RISC-V, the open-source chip architecture, through India’s DIR-V programme — letting Indian companies design specialised chips without depending on ARM licensing. If the India semiconductor strategy pivots to inference, RISC-V is the vehicle. It’s a sovereignty play as much as a technology one.

The Three-Point Plan

Chandrasekhar didn’t just diagnose; he prescribed:

• A National Semiconductor Research Institute focused on process technology, design IP and talent development — the deep research base India currently lacks.
• A ₹1,000-crore Chip Design Commercialisation Fund, modelled on the National Investment and Infrastructure Fund, to help Indian fabless companies finance development through commercial tape-out — the expensive step where designs become physical chips.
• At least two sovereign AI inference-chip programmes with guaranteed government offtake — meaning the government promises to buy the chips, de-risking the bet for whoever builds them.

Why This Debate Matters Now

The timing isn’t accidental. AI is simultaneously reshaping India’s IT industry — analysts note that large-cap IT firms are expected to grow just 2% in FY27 as clients rethink spending around AI, while nimbler midcap firms adapt faster. The whole industry is shifting from billing by headcount to outcome-based models. In that world, owning chip design capability isn’t a prestige project; it’s economic infrastructure.

Whether the government adopts this version of the India semiconductor strategy is an open question — he’s now outside the government, commenting as a policy voice. But the argument itself is hard to dismiss: in the AI supercycle, the countries that design the silicon will capture far more value than the countries that merely package it. India has the engineers. The real question for the India semiconductor strategy is whether the country will bet on them.

For students eyeing this future — chip design starts with strong fundamentals. Our CGPA to percentage calculator and GST calculator are small tools for the everyday math along the way.

Source reporting: Business Today and NewsBytes.