India has 125,000 chip designers, and not one of them designs a chip that India owns.

Niti Aayog member V.K. Saraswat said the quiet part out loud last week when he told Mint that Indian designers contribute to virtually every chip produced globally but work primarily for MNCs, mostly on subsystems and components, with very few capable of what he called “ab initio design of a complete system.”

What makes this significant is not the diagnosis — anyone in the industry knows this — but the fact that a senior government figure finally framed India’s celebrated design talent as a structural dependency rather than a national strength.

..

Because what Saraswat described is India’s IT services model — labour arbitrage without value creation — replicated inside silicon. Intel employs 14,000 engineers in Bengaluru, AMD has 6,500 across ten locations, and in February 2026 Qualcomm completed a 2nm chip tape-out entirely from its Indian centres, which is a genuine frontier achievement.

But the architecture decisions, the product strategy, and the patent filings all stay at headquarters, which is why Qualcomm filed 3,482 patent applications at the Indian Patent Office in 2022-23 while Cadence India’s own managing director acknowledged publicly that what India is missing is homegrown product companies. India provides the brains; San Diego captures the value.

The pride in “Indians design every chip” is precisely the delusion that prevents the country from asking why no Indian company designs any chip worth naming.

..

And yet a handful of startups have quietly proven that end-to-end chip design from Indian soil is an accomplished fact. Signalchip, backed by Zoho’s Sridhar Vembu, built India’s first indigenous 4G/5G modem chips — one of roughly eight companies on Earth with this capability, now deployed in Indian Army base stations.

Mindgrove Technologies shipped a commercial RISC-V microcontroller fabricated at TSMC’s 28nm node. Saankhya Labs developed production software-defined radio chipsets. ISRO’s Vikram 3201 is a flight-tested, fully indigenous space-grade processor. These are real chips with real deployments, and they validate Saraswat’s premise that the talent exists — the question is entirely about what kind of ecosystem surrounds it.

..

Which is where his diagnosis collides with budget reality. The Design Linked Incentive scheme, the only government programme directly funding chip design startups, has a total allocation of roughly ₹1,000 crore — barely 1.3% of the semiconductor mission outlay — and supports just 24 companies whose combined VC funding of ₹430 crore is less than what AMD spends in India in a single year.

ISM 2.0’s R&D allocation works out to approximately $120 million, which would not cover the cost of a single advanced SoC programme at Qualcomm or Apple. For perspective, China poured $98 billion across three Big Fund phases and built a $91 billion fabless industry from near-zero in two decades, and Taiwan spent 20 years building a deliberate pipeline from ITRI’s government-funded research to the TSMC spinoff to MediaTek.

We’re saying the scale of this challenge is many times larger than the budget we have for it.

..

Saraswat says the mission will recruit MNC designers to build indigenous capability, but the question he didn’t answer is what they’d be recruited to — because Signalchip’s founder survived his first four years on $2 million from family and friends after every VC in the country turned him away, and the DLI scheme’s per-company allocation of roughly $2.4 million is enough for one mask set at a trailing-edge node but nowhere near enough to build a globally competitive product.

There is no government procurement mandate for indigenous chips, no patient capital fund that understands decade-long hardware return horizons, and no DARPA equivalent guaranteeing first customers.

The startups have proven the concept, but until the distance between what gets said at Niti Aayog and what gets funded by the Finance Ministry narrows dramatically, Indian chip sovereignty will remain a conference-stage aspiration rather than an industrial reality.

At the World Economic Forum in Davos on January 20, 2026, Union Minister Ashwini Vaishnaw found himself in a position familiar to any Indian representative at global forums: being talked down to.

When IMF Managing Director Kristalina Georgieva casually placed India in a “second tier” of AI nations—behind America and China, lumped with capable but lesser powers—Vaishnaw pushed back with vigour.

“India is clearly in the first group,” he insisted, citing the Stanford AI Index: third globally in AI vibrancy, second in talent concentration. The country, he argued, was already “the largest supplier of AI services” and building systematically across five foundational layers: applications, models, chips, infrastructure, and energy.

Vaishnaw was doing his job, and doing it well. It would have been absurd for India’s Minister of Electronics and IT to sit on a Davos stage and enumerate his country’s capability gaps for an international audience.

At global forums, you project strength, claim your place at the table, and let the rankings speak. His citations were accurate; his framing was strategic; his confidence was appropriate. When the IMF Managing Director casually diminishes your country’s standing, you push back. This is what ministers do, and Vaishnaw did it effectively.

..

But what serves India well at Davos does not serve Indians well at home. The risk is that ministerial confidence curdles into national complacency—that we mistake projecting strength for possessing it, that we confuse a good performance with a won race.

Vaishnaw’s five-layer framework is genuinely useful for understanding India’s AI position. Examining each layer honestly, however, reveals a more complex picture than triumphalism allows.

The ₹10,372 crore ($1.25 billion) IndiaAI Mission represents serious intent; it also represents roughly what OpenAI or Anthropic burns through in six months.

Understanding where India actually stands—and what must be built—requires examining each layer against global benchmarks, with the honesty that Davos does not permit but domestic strategy demands.

..

Layer 1: Applications — India’s Genuine Strength, Within Limits

India’s claim to AI applications leadership deserves serious examination… (continue reading on Swarajya)

CONTINUE READING ON SWARAJYA

HAL was building aircraft before Brazil had an aerospace program. Now we’re assembling their jets.

In December 2025, Adani signed a deal with Brazil’s Embraer to set up India’s first commercial aircraft assembly line. Not manufacturing—assembly. India provides the facility; Brazil provides the aircraft, the intellectual property, the certification expertise, and the global credibility. After 85 years of aerospace activity, this is what arrival looks like for India.

..

Here’s the uncomfortable math. Hindustan Aeronautics Limited was founded in 1940. Embraer was established in 1969—nearly three decades later. India had the head start, the technical workforce, the institutional foundation. Brazil, emerging from military dictatorship with little more than an aeronautics institute, came from behind to become the world’s third-largest commercial aircraft manufacturer.

Embraer has delivered 9,000 aircraft to customers across 70 countries. Revenue in 2024: $6.4 billion, with a $26.3 billion order backlog. HAL? Revenue of $3.6 billion, with 95% coming from a single customer—India’s armed forces. Exports constitute 1% of sales. HAL has never successfully developed a commercial aircraft that any airline outside the subcontinent would purchase.

..

The divergence isn’t about talent. India’s space program reaches Mars. Its software engineers build products used worldwide. This is about what happens when you give a company a captive market and call it a strategy.

HAL was nationalised in 1964 and became the sole supplier to India’s armed forces. When the Air Force needed fighters, HAL assembled Russian MiGs under licence. When it needed trainers, HAL built British Hawks. The company has assembled 660 MiG-21s, 272 Su-30MKIs, and 82 Hawk trainers. Impressive numbers that mask a troubling reality: licensed production transfers assembly skills, not design capability. The institutional incentive was to fulfil government orders, not to develop products that could compete internationally.

The Tejas fighter program, sanctioned in 1983 to replace ageing MiG-21s, achieved initial operational capability in 2015. A 32-year development cycle. Meanwhile, Embraer was exporting aircraft within six years of founding.

..

Brazil made five choices India refused to make.

First, export focus—by 1989, half of Embraer’s revenue came from international sales, forcing the company to meet global quality standards.

Second, privatisation in 1994, which introduced profit-driven management and killed bureaucratic inertia.

Third, risk-sharing partnerships that distributed development costs among global suppliers.

Fourth, niche focus—Embraer dominated the 70-150 seat regional jet segment that Boeing and Airbus ignored, rather than spreading resources across fighters, helicopters, trainers, and transports simultaneously.

Fifth, reinvestment over dividends—while HAL paid cash to the finance ministry, Embraer and its partners invested in design capability.

The result? When Indian airlines order 1,800 aircraft over the next two decades, they’re ordering from Boeing and Airbus. None from HAL.

..

The Adani-Embraer deal isn’t a failure. It’s a diagnosis. Final assembly represents 15-20% of an aircraft’s value. The core intellectual property—aerodynamic design, fly-by-wire systems, engine integration—stays in São José dos Campos. India gains jobs but not the technological capability that would enable indigenous development.

India’s engineering talent was never the constraint.

The constraint was policy architecture that treated aerospace as a defence procurement programme rather than a commercial industry requiring market discipline.

Brazil, starting later but choosing differently, built a global champion. India, with every advantage of time and talent, built a government contractor. The 30-year head start didn’t just disappear. We spent it.

This article is based on “Adani-Embraer Deal Raises the Question: What Could HAL Have Been?“ by Raghavan S Rao, published in Swarajya.

A fab without chemicals is a ₹91,000 crore paperweight.

India’s semiconductor mission gets breathless coverage every time a foundation stone drops. The Tata fab in Dholera. Micron in Sanand. Ministerial photo-ops and billion-dollar announcements. But here’s what nobody’s talking about: a chip factory is a ravenous chemical organism. It inhales hundreds of distinct inputs—acids, gases, solvents—and if even one is contaminated by a few parts per billion, the entire production line crashes to zero yield. India currently imports nearly 100% of these materials.

The ₹76,000 crore India Semiconductor Mission corpus? Already committed. The fabs are coming. But the invisible infrastructure that keeps them alive—the ultra-pure molecules flowing through fluoropolymer-lined pipes—that’s where sovereignty will actually be decided.

..

Here’s India’s absurd position: the country is a global powerhouse in industrial chemicals, exporting billions in fluorochemicals and solvents annually. Yet it remains a net importer of the exact same chemicals in electronic-grade purity. The difference between 99.9% pure (industrial) and 99.9999999% pure (semiconductor-grade, called “9N”) isn’t a decimal point. It’s a civilizational capability gap.

Take hydrofluoric acid—the workhorse for etching silicon. India makes industrial HF. But the Grade 5 purity required for advanced chips? Historically imported from Japan’s Stella Chemifa and Morita. Tanfac Industries is now racing to upgrade its Tamil Nadu plant to hit parts-per-trillion impurity levels. Gujarat Fluorochemicals is going further—buying fluorspar mines in Morocco to control the raw material for the fluoropolymers that line every pipe and valve in a fab. Steel corrodes in these environments. Only ultra-pure PFA and PVDF plastics survive.

..

A fab also breathes. Massive quantities of nitrogen, argon, and hazardous specialty gases like silane and nitrogen trifluoride. If the nitrogen supply hiccups, oxidation furnaces fail. Inox Air Products has locked in the first-mover advantage with a ₹500 crore Electronic Specialty Gas Hub in Dholera—designed to localize storage, mixing, and purification of 40+ gases currently shipped in cylinders from overseas.

Then there’s water. A single fab drinks at the scale of a small city. In drought-prone Gujarat, this means mandatory desalination and zero-liquid-discharge systems. The Dholera project requires 100% hazardous wastewater treatment. Companies like Thermax and VA Tech WABAG are deploying membrane bioreactors to recycle 90% of the water—because the alternative is political disaster in a water-stressed state.

..

The logistics are equally invisible and equally critical. You can’t move ultra-pure acids in standard steel tankers—the metal leaches into the fluid, ruining it. You need ISO tanks lined with fluoropolymers. Silane ignites spontaneously on contact with air. Specialized carriers like Stolt Tank Containers are building the “invisible infrastructure” to move these cargoes from Mundra port to Sanand cleanrooms without killing anyone or contaminating anything.

Meanwhile, China’s Big Fund III just deployed $47.5 billion to flood global markets with subsidized legacy chips. If India can’t stand up domestic chemical supply chains fast enough, its shiny new fabs will be competing against predatory pricing with one hand tied behind their back—still dependent on imported molecules that could be cut off tomorrow.

..

The fabs will get built. The ribbon-cutting ceremonies will happen. But semiconductor sovereignty isn’t measured in foundation stones. It’s measured in nine-nines purity, in fluoropolymer-lined pipes, in specialty gas hubs, in desalination plants.

India’s chip ambitions are real. The question is whether the invisible war—the one fought in chemistry labs and logistics networks—will be won in time. The structures rising in Dholera and Sanand are trenches. The chemicals flowing through them are ammunition. And right now, India is still importing most of its bullets.

Credit: Based on reporting from Swarajya

ISRO launched four rockets this year. One failed. Critics are furious about the low count.

Here’s what they’re missing: this isn’t incompetence. It’s strategy. India’s space agency is deliberately stepping back from the launch business—so that private industry can step in. The low numbers aren’t a bug. They’re a feature of a transition that’s been brewing for years.

..

Let’s get the facts straight first. Of ISRO’s four 2025 launches, three succeeded—the January GSLV flight delivering a NavIC navigation satellite, the July NISAR mission with NASA, and the November LVM3 launch placing a communication satellite in orbit. The lone failure was the May PSLV mission, which aborted mid-flight due to a chamber-pressure anomaly in the third stage. Not one of these was a commercial mission.

The target was 30 launches between January 2024 and March 2025. They managed around eight. That’s roughly a 73% shortfall. So yes, the criticism about cadence is valid. But here’s where context matters.

..

ISRO’s rockets—PSLV, GSLV—were never designed for mass production. They’re hand-assembled vehicles from decades-old designs. You can’t roll them off a production line like SpaceX does with Falcon 9. At roughly $2,600 per kilogram to orbit, Falcon 9 is so cheap that even Indian startups choose to fly American. ISRO’s former chairman admitted it bluntly: “Rockets are built and kept in stock, but not finding customers.”

The agency has capacity three times higher than demand. That’s not a manufacturing problem—it’s a market problem. And ISRO’s response has been strategic withdrawal rather than futile competition.

..

Here’s where the real shift is happening. ISRO has fundamentally changed how it operates. Earlier, satellites were built speculatively—launch first, find users later. Now, every launch requires a committed customer before ISRO proceeds. This kills vanity metrics but eliminates waste.

Meanwhile, the agency is redirecting its best minds toward what only a national space agency can do: Gaganyaan (India’s first crewed mission, targeting 2028), Chandrayaan-4 (lunar sample return), the Bharatiya Antariksh Station, and landing an Indian on the Moon by 2040. The unglamorous work of routine satellite launches? That’s being handed to the 350-plus space startups now operating in India.

Prime Minister Modi has set the target: 50 launches per year within five years. One rocket every week. That’s only possible if private industry industrialises production of LVM3 and SSLV—ISRO’s newer, more production-friendly vehicles. Sixty-plus LVM3 rockets are already approved for private-sector manufacturing.

..

The next 12 months tell the real story. Multiple missions are lined up through March 2026. The Rs 27,000-crore SBS-3 surveillance constellation—52 satellites over five years—kicks off soon after. A new spaceport at Kulasekarapattinam will add capacity for 20-25 smallsat launches annually.

ISRO’s quiet year isn’t decline. It’s the lull before transformation. The agency that built India’s space capabilities from scratch is now betting that stepping aside will build something bigger: an entire industry.

Whether that bet pays off depends on what happens next—not on what happened this year.

Credit: Based on reporting by Karan Kamble | Swarajya

Stanford postdoc. Six months waiting. Zero acknowledgement. That’s how India welcomes its returning scientists.

A young researcher applied for a faculty position at an elite Indian institute after her Stanford stint. She expected tough competition. What she got instead was silence—no interview, no rejection, no explanation. Her application simply vanished.

“It goes into a black hole,” says a senior faculty member familiar with the case. “The entire process takes several months, and it’s very bureaucratic. They don’t even reply to the applicant’s email. If you’re not selected, you’re not told why.”

..

Here’s the thing: this is happening at the worst possible time.

Trump’s H-1B policy just added a $100,000 fee increase. Indians hold over 70% of all H-1B visas. The uncertainty has been building since 2017, pushing researchers to reconsider long-term US careers. Meanwhile, Europe is aggressively poaching talent—Germany’s Skilled Immigration Act, France’s Talent Passport.

India has a rare window. About 550 researchers returned under the Ramanujan Fellowship this decade. The government is designing new schemes with setup grants for labs. The DST Secretary told media they’re planning programmes to attract scientists “in the current geopolitical situation.”

The talent is finally flowing back. And Indian academia is fumbling the catch.

..

The problems run deep. At one top institute, a returnee’s application wasn’t processed due to reservation category mismatches—”departments fighting over general slots, almost no general hires,” according to someone who knows both sides.

But reservation isn’t the core issue. Faculty posts are vacant across all categories. IIT Kanpur needs 1,000 faculty for its 10,000 students. It has 600. A March 2025 parliamentary report revealed 56% of professor positions at IITs, IIMs, NITs, and IISERs sit empty.

The real rot is cultural. “What matters, honestly speaking, is which group or lobby you belong to,” says an IIT professor who also teaches at a US university. “If you’re not from these lobbies, it’s very difficult to enter or survive.” Certain communities hire their own. Faculty members recruit their former students. State universities demand bribes for permanent positions. Anna University uncovered 2,000 cases of fake faculty appointments in one academic year—one person employed by 32 colleges simultaneously.

..

Even when researchers get hired, the ecosystem can’t support them.

Western professors have dedicated labs. Indian scientists share generic, all-purpose facilities. Setting up a proper research lab costs ₹10-15 crore, plus running costs, equipment, assistants, sensors, raw materials. “The institute has to be very liberal for someone used to abundance in the West to think of coming back,” says a senior engineer connected to aerospace and defence R&D.

Then there’s academic freedom. PhD admissions run through reservation quotas that don’t match project needs. A professor might want an excellent SC candidate, but the available slot is for OBC. “When they don’t get that academic freedom, they’re not going to enter the IIT system.”

The infrastructure gap means even fixing hiring won’t fix outcomes. Scientists either won’t come, or won’t stay.

..

India scrapped 300 science awards in 2022 after the same corruption plagued that system. The reset worked. Similar thinking could apply to hiring: transparency mandates, external reviewers for all appointments, digital verification to catch ghost faculty.

The ANRF could offer repatriation grants—not fellowships, but ₹15-20 crore capital grants for returning scientists to build labs exactly as they envision. Reports suggest the government is considering setup grants in new re-entry schemes.

But none of this matters without enforcement. The Anna University scandal only broke because a civil society group exposed it. A national academic integrity commission with suo motu powers could build vigilance into the system instead of waiting for whistleblowers.

The talent is knocking. The geopolitics align. Indian academia needs to decide whether it wants to answer the door—or keep pretending nobody’s home.

Credit: Based on reporting by Karan Kamble | Read full article on Swarajya

For centuries, it existed only as footnotes. Scattered citations in medieval commentaries. A title without a body. Mahāsalilam—the twenty-fourth Aṅga of the Vṛddhagārgīya Jyotiṣa—was India’s oldest astronomical text. And it was lost.

Until Prof. R. N. Iyengar decided to find it.

..

Here’s what that quest looked like: Eleven manuscripts. Spread across India and abroad. Some incomplete. Some corrupted by centuries of copying errors. Each fragment preserved a piece of a 3,800-year-old conversation between two sages—Viśvāmitra asking questions about the cosmos, Vṛddha Garga answering with precision that would make a modern physicist pause.

Iyengar spent twenty years collating them. Comparing variant readings. Decoding archaic Sanskrit terms that had no modern equivalents. Cross-checking astronomical data against verifiable celestial events from 1800 BCE. What emerged was not just a book—it was evidence.

Evidence that India’s astronomical tradition didn’t borrow from Babylon or Greece. It predated them.

..

The timing of the book’s launch tells you everything. On December 28, 2024, Dr. S. Somanath—Chairman of ISRO—released the critical edition at Bengaluru’s Mythic Society. Not a historian. Not an archaeologist. The man who runs India’s space programme.

Because Mahāsalilam isn’t archaeology. It’s a continuum.

The text describes 108 celestial bodies—seven visible planets and 101 comets. It quantifies brightness and motion. It maps the six-monthly lunar eclipse cycle. It outlines the five-year Yuga calendar that tracks solstices with precision. These aren’t myths. These are observations recorded before the Greeks started stargazing.

Look at the numbers: The summer solstice aligned with the Maghā constellation. That places the text in the Maghādī era—1800 to 1600 BCE. That’s older than Lagadha’s Vedāṅga-Jyotiṣa, which Western scholars grudgingly accept as India’s earliest astronomical work. Mahāsalilam predates it by centuries.

..

Now here’s what nobody talks about. Iyengar isn’t just a Sanskritist. He’s a trained engineer and seismologist from the Indian Institute of Science. Which means he approached this like a forensic investigation, not a literary exercise.

Every claim in the text—comet paths, eclipse timings, polar star identifications—was cross-checked against computational astronomy. The result? The data holds. Vṛddha Garga wasn’t theorizing. He was recording what he saw. The text even identifies Thuban as the pole star, which was astronomically accurate for that era.

..

The book itself is dense—321 pages of critical edition, translation, and commentary. Published by the Central Sanskrit University. Padma Shri Prof. M. D. Srinivas wrote the foreword, positioning Mahāsalilam in the arc from Vedic astronomy to the Siddhāntas. Every section includes notes explaining technical terms in both Vedic and modern astronomical contexts.

But the real story isn’t the book’s contents. It’s what the book represents: proof that India’s curiosity about the cosmos is unbroken. From identifying Abhaya Dhruva as the pole star 3,800 years ago to navigating satellites today, the lineage is clear.

ISRO’s chief didn’t release this text for nostalgia. He released it because the same civilisation that mapped solstices before writing was widespread now launches lunar missions. That’s not coincidence. That’s continuity.

And it took one scholar twenty years to prove it.

Original Reporting: Before Space Science, There Was Mahāsalilam: Rediscovery Of India’s Oldest Astronomical Text

556 days. That’s how long it took Anduril to go from clean-sheet design to first flight of the YFQ-44A Fury.

The Fury is Anduril’s entry in the Collaborative Combat Aircraft (CCA) program—the US Air Force’s bet on autonomous drones that fly alongside manned fighters as “loyal wingmen.” These aren’t the Predator drones you’re thinking of. CCAs are jet-powered, semi-autonomous aircraft designed to operate in contested airspace. They carry weapons, jam enemy radar, and can make targeting decisions using AI. The goal: give each F-35 or F-22 pilot a team of robotic wingmen that multiply combat power without risking more human lives.

The US Air Force wants 1,000 of these CCAs by the early 2030s to counter China’s numerical advantage in fighters. The program is worth at least $30 billion, potentially $180 billion over its lifecycle.

And Anduril—a seven-year-old startup founded by Palmer Luckey—just beat Boeing, Lockheed Martin, and Northrop Grumman to build it. The other company tasked with building CCAs is General Atomics.

..

The F-35 took 23 years from contract to full-rate production. The F-22 needed 6 years just to fly. The B-21 stealth bomber? 8 years. And here’s a 32-year-old college dropout—Palmer Luckey, the kid who built Oculus in his parents’ garage—who just built a semi-autonomous fighter drone in 18 months.

This isn’t incremental improvement. This is proof that the entire US defense acquisition system is fundamentally, catastrophically broken.

The timeline doesn’t lie. Contract award: April 24, 2024. First flight: October 31, 2025. That’s 18 months from “go” to “airborne.” Meanwhile, Boeing spent a decade and billions making the KC-46 tanker that still leaks fuel. Lockheed Martin has been iterating the F-35 since 2001 and it’s still having software problems in 2025.

..

What’s Anduril’s secret? It’s not magic—it’s venture capital and speed over bureaucracy.

Traditional US defense acquisition is sequential, rigid, and designed for political optics, not warfighting. Requirements get locked in for years. Every milestone requires congressional review. The entire process incentivizes contractors to drag things out because cost-plus contracts reward spending, not delivering.

Anduril flipped the script. They built with private capital first—$75 million of their own money matched against the Air Force’s $14.3 million. They iterated rapidly. They failed fast. And most critically, they didn’t wait for the Pentagon to tell them what to build.

The numbers are even more brutal when you zoom out. Anduril’s Arsenal-1 facility—a 5 million square foot factory in Ohio—cost nearly $1 billion. That’s private investment, not taxpayer money. It opens in July 2026 and will produce “tens of thousands” of systems at scale. For context, Lockheed Martin’s F-35 production line makes about 150 fighters per year. Anduril’s factory will produce more aircraft in a year than Fort Worth produces in five.

..

And they’re still getting faster. During the prototype phase, Anduril doubled their manufacturing speed through continuous optimization. Their ArsenalOS software platform integrates the entire production process—no siloed systems, no manual handoffs, no defense contractor inefficiency.

Air Force Secretary Troy Meink called it proof that “competition drives innovation and accelerates delivery.” But let’s be honest about what he’s really saying: the traditional US defense primes have been moving at glacial speed because they could. Because there was no competition. Because cost-plus contracts made delay profitable.

When General Atomics and Anduril both got their CCAs flying in under two years, they proved something the Pentagon has suspected for decades: the acquisition process inflates timelines by 5-10x artificially. The bureaucracy isn’t protecting quality—it’s protecting incumbent contractors.

..

Jason Levin, Anduril’s VP, put it plainly: “Faster than any major fighter aircraft program in recent history.” That’s not marketing speak. That’s a threat to every US defense prime.

Here’s the uncomfortable question: If a startup with venture backing can move 12x faster than Lockheed Martin, what does that say about every US defense program of the last 50 years?

Analysis based on public reporting on Anduril CCA program, Mitchell Institute wargames, and US Air Force acquisition strategy

China lost track of two missiles in 1996. America’s been paying for that ever since.

The story: China fired three missiles toward Taiwan as a warning during a exercise. Two disappeared mid-flight. The US had allegedly cut GPS access. Beijing’s military couldn’t track its own weapons. Chinese documents still call it an “unforgettable humiliation.”

Fast forward to 2025. China operates 56 navigation satellites. America has 31.

..

In 165 countries, capital cities see more Chinese satellites than American ones. Your delivery app in Ethiopia? Chinese satellites. Port management in Pakistan? Chinese. That “free” navigation spreading across Africa and Southeast Asia? All Beijing.

The system is BeiDou. It generates $79.9 billion annually and runs on 288 million smartphones. It all traces back to 1996 when China’s missiles went dark.

..

The thing about humiliation is it focuses the mind.

China started building BeiDou in 1994. First satellites launched in 2000. But here’s where it gets interesting—China didn’t just copy GPS. They studied what makes satellite navigation vulnerable and built backups for the backups.

GPS lives entirely in space. Jam the signal, you’re done.

China built three layers: satellites, ground-based eLoran towers, and fiber-linked timing networks. You can knock out one. Knock out two. But all three? That’s a different problem.

..

Think about a Taiwan conflict. China jams GPS across the region. US forces, Japanese ships, allied aircraft—all blind. Chinese missiles keep firing because eLoran towers don’t care about satellite jamming.

The asymmetry is brutal.

Meanwhile, America’s GPS satellites average 13 years old. The oldest has been up there 27 years—since Blockbuster was booming. GPS 3 offers 1-3 meter accuracy. BeiDou does centimeter-level positioning.

The President’s Advisory Board on GPS put it bluntly: “GPS’s capabilities are now substantially inferior to those of China’s BeiDou.”

..

China has 10 times more monitoring stations globally than GPS. Those stations collect data on user locations, movement patterns, infrastructure dependencies. In the Global South where BeiDou provides the most accurate positioning, China has leverage.

Cut off navigation service? That’s economic warfare without firing a shot.

BeiDou’s two-way messaging makes it worse. Unlike GPS, these satellites receive data from users. Search and rescue, sure. But also surveillance. Real-time intelligence from every device.

..

The US is finally waking up. GPS 3F satellites are coming. Proposals for terrestrial backups. Plans for better authentication.

Too late.

China’s embedded in 165 countries. The person in Addis Ababa ordering food has no idea Chinese satellites make it work. They just know it’s better than alternatives.

The long game Beijing’s played since 1996.

..

Thirty years later, China doesn’t just have its own GPS. It has a better one with more satellites, more stations, more features. More countries depend on it than the system it replaced.

The humiliation didn’t weaken China’s resolve. It crystallized it.

And now the US is playing catch-up.

India’s launching yet another big government scheme focused on R&D.

You’ve heard this before.

But here’s what’s different: the Department of Science and Technology just got permission to ignore the rules that kill every other innovation initiative.

The Rs 1 lakh crore Research, Development and Innovation Fund isn’t just bigger than past schemes—it’s structurally different. The difference comes down to three letters: GFR.

..

General Financial Rules govern how government departments spend money. They exist for accountability and transparency. They also make innovation nearly impossible.

Here’s what GFR does in practice: You need a quantum computing specialist. GFR says advertise, form committee, interview, offer government wages. Six months minimum. The specialist takes a private job at three times the salary.

You need rapid prototyping. GFR says lowest bidder wins. The capable firm quotes realistic costs. An inexperienced firm quotes 30% less. You’re stuck with the cheaper option and a failed project.

You want fast decisions. GFR requires five approval layers with full documentation. By the time you get clearance, the opportunity’s gone.

This is how Indian government innovation has worked for decades.

..

After two months of negotiations, DST secured a GFR exemption. Not full exemption—but enough to matter.

The fund can now pay market salaries, hire quickly, contract based on capability not just cost, and move capital at private-sector speed.

These exemptions are granted in “rarest of rare situations,” according to policy analysts. Most departments never get them. DST fought for two months and won.

..

Here’s why that matters: India’s tried big R&D schemes before. Every time, they moved at government speed—which is to say, barely at all.

DRDO has Rs 25,000 crore annually. Most goes to salaries and operations. Compare that to America’s DARPA, which operates with similar budget but private-sector agility and has seeded the internet, GPS, and stealth technology.

CSIR spends its entire Rs 6,000 crore keeping labs running. Research happens, but commercialization? Rare.

The problem isn’t money or intent. It’s friction. You can’t compete for talent at government wages. You can’t make fast decisions through five-layer approvals. You can’t build innovation with lowest-bidder contracts.

The exemption removes that friction.

..

What makes this interesting: the fund has structural permission to operate differently.

It’s housed in the Anusandhan National Research Foundation, which has parliamentary backing and a 50-year timeline. Second-level fund managers will be professional VCs and DFIs—not government committees. Companies must co-invest 50%.

But none of that works if hiring takes six months and every decision needs bureaucratic sign-off. The GFR exemption makes the architecture functional rather than theoretical.

..

What to watch: DST allocated Rs 20,000 crore for 2025-26. First disbursements expected mid-2026.

Key signs over the next 18 months: How quickly does DST hire specialists? What caliber of fund managers get selected? How fast do first investments happen?

The exemption creates possibility. Execution determines reality.

..

Here’s the honest part: this could still fail. Having permission to move fast doesn’t mean DST will actually move fast. Decades of institutional culture don’t change overnight.

But—and this matters—the tools exist. Past initiatives failed because the structure made success impossible. This one has structural permission to succeed.

..

Why care?

Because India spends 0.8% of GDP on R&D while South Korea spends 4%+.

Because at current pace, it takes 70 years to reach even 25% of US per-capita income.

This fund is India’s bet on building innovation capability. The GFR exemption is what makes the bet credible rather than symbolic.

This is the first major Indian R&D initiative with structural permission to operate at market speed. Whether that permission translates to actual speed depends on execution over the next 2-3 years.

The opportunity exists. Now we see if DST uses it.

Original reporting: Inside Modi Government’s Rs 1 Lakh Crore Plan To Turn Around R&D In India

In 2020, India did something that looked insane. It cancelled global 4G tenders and decided to build its own network. And spent Rs 37,000 crores on it.

While the world was deploying 5G.

The math seemed backwards. Spend years developing “outdated” technology while competitors raced ahead. Indian telcos would have to learn maintenance while fighting global players with better infrastructure.

But here’s what that decision actually bought.

..

When you depend on Ericsson, Nokia, or Huawei for your network, you’re not just buying equipment. You’re outsourcing control.

They manage the upgrades. They handle the security patches. They see the data flows.

And when sanctions hit or geopolitical tensions spike, countries discover they’ve built critical infrastructure on rented ground.

Remember what happened with Huawei? Countries had to rip out billions in equipment overnight. Not because the tech failed—because trust did.

..

India’s telecom imports were Rs 1.5 lakh crore in FY23-24. That figure isn’t just money. It’s strategic vulnerability.

Because by 2025, your network isn’t just for phone calls. It’s the backbone for everything that matters. Military systems run on it. Healthcare data flows through it. AI and cloud computing depend on it.

When that infrastructure is foreign-controlled, you don’t have a telecom policy. You have a dependency problem.

..

So India built its own stack.

C-DOT designed the core. Tejas Networks and TCS developed the radio access network. The entire supply chain—from design to deployment—became indigenous.

26,000+ sites went live in 18 months using homegrown technology.

Not because it was patriotic. Because it was the only way to own the encryption, control the upgrades, and manage who sees what data.

..

Yes, India’s late to 5G.

Yes, Indian telcos are learning maintenance while competing globally.

But here’s what changed: when 6G applications hit critical sectors—defense, healthcare, agriculture—India owns its infrastructure. The encryption protocols are Indian. The security audits don’t need foreign approval. The upgrade path isn’t held hostage to vendor relationships.

..

Other countries in Asia, Africa, and Latin America are watching this closely.

They face the same choice: cheap dependence or expensive sovereignty.

India’s bet is simple. By the time 5G and 6G matter for sectors where control is everything, being two years late but independent beats being on-time but owned.

..

The question was never whether India could afford to build its own network.

It’s whether India could afford not to.

Because the real cost of foreign infrastructure isn’t measured in crores. It’s measured in who controls what when it matters most.

And that’s not for sale at any price.

X Corp walked into Karnataka High Court demanding special treatment.

The company monetizes 98 crore Indian users—collecting their data, selling their attention, profiting from their engagement. But when India asked them to remove illegal content? Suddenly they’re constitutional scholars.

Their argument: Section 79(3)(b) of the IT Act and the Sahyog Portal—the government’s content-blocking notification system—are unconstitutional. Only Section 69A with its slower procedures should apply.

Translation: We want to profit from India’s users while deciding for ourselves what Indian law means.

Here’s the part X Corp didn’t mention: They’re a foreign corporation. They have zero constitutional rights in India. But they wanted Indian courts to declare Indian laws unconstitutional anyway.

..

X Corp’s Business Model Required Looking the Other Way

Solicitor General Tushar Mehta destroyed X Corp’s argument with simple math.

India’s internet users grew from 25 crore in 2014 to 98 crore today. Data consumption: over 20GB per user per month. That’s X Corp’s revenue stream—more users, more data, more money.

But here’s what comes with those numbers: COVID misinformation that killed people. Lynch mobs organized through viral posts. Communal violence triggered by fake news. Financial scams spreading at algorithmic speed.

X Corp wanted safe harbor protection—legal immunity for all that content. The government’s response was brutal: You don’t get immunity if you refuse to act on illegal content.

The trick is simple: Section 79(3)(b) isn’t about censorship. It’s about liability. When the government notifies you that content is illegal, you remove it or you lose your protection.

X Corp’s real complaint? The Sahyog Portal makes this process too fast. They wanted the slower Section 69A procedures—more time to profit from viral content before having to take it down.

And this is where it gets worse: X Corp can’t even claim free speech rights. Article 19(1)(a) protects Indian citizens, not foreign corporations. They were arguing Indian constitutional law while having zero constitutional standing.

..

The Court Rejected X Corp’s Self-Serving Arguments Completely

Karnataka High Court didn’t just rule against X Corp. It dismantled their entire position.

Both Section 69A and Section 79(3)(b) are valid. The Sahyog Portal stays. X Corp’s challenge? Dismissed.

The Court’s verdict cut through X Corp’s rhetoric: “The digital public square, while vital for free expression, cannot operate unchecked, as unregulated content poses real risks to public order and citizens’ rights.”

You know what happens next? X Corp tries to hide behind algorithms. “We can’t control what the algorithm shows.” The Court killed that excuse too. Humans write algorithms. Humans program the rules. X Corp can’t claim their automated system is neutral when they designed it to maximize engagement—even if that engagement spreads misinformation.

Think about it: X Corp cited Shreya Singhal, a 2015 judgment, as if nothing changed. The Court pointed out the obvious—India had 25 crore internet users then. It has 98 crore now. The internet was slower, less dangerous, less weaponized. Legal frameworks evolve because reality changed.

X Corp’s entire defense was “let us self-regulate.” The Court looked at COVID misinformation, communal violence, and organized crime spreading on these platforms and said: Self-regulation failed. The experiment is over.

..

What X Corp Really Lost

The right to cherry-pick which Indian laws to follow.

The ability to profit from illegal content while claiming ignorance.

The fantasy that foreign corporations can dictate terms to sovereign nations.

But wait. X Corp still operates in India. Still has 98 crore users. Still makes money from their data.

The difference: They now have to actually respond when the government notifies them of illegal content. They have to act or lose their legal immunity.

That’s what this whole fight was about. X Corp wanted to collect Indian data, monetize Indian attention, profit from Indian engagement—while being exempt from Indian accountability.

The math is brutal: You can’t have it both ways.

So who wins? Not platforms claiming they’re above local laws. The Court chose what every sovereign nation eventually chooses—foreign companies can operate here, but they follow our rules.

X Corp can keep running its digital town square in India. They just learned what every vendor in every physical town square already knows: the square isn’t yours. You’re just renting space in it.

And if you’re profiting from the crowd, you’re responsible for the chaos.

Based on legal analysis from Swarajya’s coverage of X Corp v. Union of India.

Picture this: A contractor finishes building a road in Bengaluru. Five years pass. The drains haven’t been cleaned once. Not because they’re clogged—because they don’t need to be.

Meanwhile, three streets away, another road gets dug up for the fourth time this year.

Same city. Same rainfall. Completely different outcomes.

..

What separates these two realities?

One word: TenderSURE.

In 2011, a Bengaluru-based not-for-profit called Jana Urban Space Foundation wrote a set of guidelines that solved nearly every urban infrastructure problem plaguing Indian cities. Traffic chaos. Broken footpaths. Flooding. Spaghetti cables hanging tree-to-tree. Women unable to walk safely.

Not separately. Simultaneously.

A decade later, Bengaluru has 100 kilometers of these roads. They work beautifully.

The other 14,000 kilometers? Still stuck in the old cycle.

..

Here’s the part that stings: following TenderSURE is optional.

“There are no mandatory guidelines for urban roads in India,” says Nithya Ramesh, Director of Urban Design at Jana Urban Space Foundation. “Engineers can choose to follow bits and pieces as they deem fit. This becomes a major issue when it comes to uniformity and sustainability.”

Think about that for a second.

Building height limits? Mandatory. Violate them, face consequences.

Road construction standards that prevent endless re-digging and save crores? Completely optional.

..

So what exactly is TenderSURE?

Most people think it’s about footpaths. It’s not.

TenderSURE is systems thinking applied to infrastructure. Here’s what makes it revolutionary:

💡 All eight utilities—storm water drain, sewage, power, telecom, gas, street light, surveillance—go under the footpaths on both sides

💡 Inspection chambers every 20-30 meters allow maintenance without digging up roads

💡 Single contractor accountability replaces the chaos of 20 different contractors digging at 20 different times

💡 Operations and maintenance clauses enforce quality—contractors pay penalties if roads break

💡 Equitable right-of-way distribution ensures pedestrians, cyclists, and vehicles all have designated space

It’s not a footpath solution. It’s an ecosystem redesign.

..

The results? Staggering.

When Jana Urban Space evaluated TenderSURE roads after a decade, the numbers told a complete story:

📊 228% increase in pedestrian traffic
📊 113% more women using the roads
📊 Near-zero maintenance requirements
📊 Drains that haven’t needed cleaning in 3-5 years

One contractor told Ramesh he hadn’t cleaned the drain even once during the entire maintenance period.

That’s not luck. That’s design.

..

But here’s what the data really reveals: these aren’t separate victories.

More women don’t walk because there are footpaths. They walk because:

• Footpaths are continuous and unbroken

• Utilities don’t require constant digging that breaks walkways

• Proper lighting makes routes safer

• Bollards prevent two-wheelers from invading pedestrian space

• The entire system works together

The 228% pedestrian surge isn’t a footpath story. It’s a proof point that when you fix infrastructure systemically, human behavior changes.

..

So why isn’t everyone doing this?

Three reasons, all cultural.

First: India’s L1 bidding system.

Contracts go to whoever bids lowest, regardless of technical competence. TenderSURE flips this—stringent qualification criteria mean only experienced contractors can even apply. Sometimes tenders have to be reissued because too few qualify.

“This in my opinion is a good sign,” says Ramesh, “because it is a sure shot way of getting the best quality.”

Second: The vehicle-first philosophy.

“When I work with city engineers across India,” Ramesh explains, “their focus, even for urban roads, is catered to moving traffic like it’s a highway. We are always looking at moving vehicular transport, rather than looking at the road as a conduit for movement and public space.”

Traffic police resist footpaths because they slow down cars. Electrical contractors resist underground utilities because overhead is easier. The entire system prioritizes vehicle speed over everything else.

Third: Bureaucratic inertia.

When TenderSURE was first proposed, a BBMP consultant wrote that it would cause “urban decay.” Only after the first roads proved themselves did attitudes shift.

Now the same pattern repeats in Uttar Pradesh, Odisha, Assam. Initial resistance. Then results. Then adoption.

But adoption takes years. And it’s never mandatory.

..

Meanwhile, the evidence keeps mounting.

Bengaluru’s Vittal Mallya Road—the first TenderSURE road built outside UB City—hasn’t been dug up once in a decade.

Church Street, St. Marks Road, Residency Road: all holding strong.

The Smart Cities project added another 30 kilometers using TenderSURE guidelines.

Uttar Pradesh is building 200 kilometers, rebranded as “CM GRIDS roads.”

Hubli-Dharwad: 20 kilometers.

Shillong: 18 kilometers.

Even Union Minister Hardeep Singh Puri visited Bengaluru to study the roads.

The model works. The proof exists. The expansion is happening.

But it’s happening city by city, project by project, with engineers choosing to adopt it rather than being required to.

..

Compare this to how other countries handle infrastructure standards.

In the United States, if a sewage department digs up a road, they sign a moratorium guaranteeing no repairs needed for five years. If something goes wrong because of poor construction, contractors pay penalties multiple times the repair cost.

Standards exist. Enforcement exists. Consequences exist.

Singapore and Thailand don’t prioritize vehicular mobility over walkability. They build for public transport and pedestrian flow first.

They view roads as public space for all kinds of movement—not just highways that happen to run through cities.

..

India, meanwhile, keeps digging.

The BBMP’s traditional practice: break a road into 20 parts, allocate to 20 contractors through 20 tenders. One does footpaths. Another does drains. Another does lighting. Another does the road itself.

They work at different times. Dig up each other’s work. Nobody’s accountable for the whole.

Bengaluru stays under construction. Budgets balloon. Citizens suffer.

And the solution sits there, proven, documented, expandable.

Just not mandatory.

..

Ramesh’s assessment is stark: “If we don’t do it, we will die.”

“It is becoming impossible to live in Bengaluru. Why will we not change and choose to move towards walkability? Delhi and Bombay are places where it’s becoming very difficult to even breathe. Bangalore is catching up.”

The stakes aren’t theoretical. They’re existential.

Indian cities face a choice: continue treating symptoms separately (dig roads for drainage, build flyovers for traffic, install cables overhead) or adopt systems thinking that solves interconnected problems simultaneously.

..

The question isn’t whether TenderSURE works.

That debate ended a decade ago with 100 kilometers of proof.

The question is: how many more years of chaos will it take before solving urban problems well becomes the law, not a choice?

Because right now, Bengaluru has the answer.

It’s just not mandatory yet.

This analysis draws from detailed reporting by Swarajya’s Adithi Gurkar, who spoke with Nithya Ramesh and the Jana Urban Space Foundation team about TenderSURE’s decade-long journey.

Imagine the beating heart of every factory, every production line, every manufacturing miracle you’ve ever witnessed.

It’s not the workers. It’s not the raw materials. It’s not even the brilliant engineers designing the products.

It’s the machines that make the machines.

Here’s the uncomfortable reality most people miss: While India celebrated its software supremacy and services dominance, it quietly surrendered one of the most critical pillars of industrial power to other nations.

And for decades, India forgot this fundamental truth while the world moved ahead. But something extraordinary is happening in workshops from Haridwar to Bengaluru that could change everything we know about India’s industrial future.

..

The missing link that no one talks about

Every manufacturing superpower has one thing in common.

Germany’s legendary Mittelstand. Japan’s precision machining excellence. China’s scaled domestic machine tool empire.

They all treated their machine-making capability like a national treasure.

India? We skipped this chapter entirely.

⭐ Manufacturing peaked at just 16-17% of GDP in India vs 25-30% in typical industrializing nations

⭐ State-led giants like Hindustan Machine Tools stagnated for decades after the 1950s

⭐ High-precision tools were overwhelmingly imported, leaving a hollow spot in India’s industrial ladder

The numbers tell a story most economists won’t admit: India leaped directly from an agrarian economy to a services-led powerhouse after Y2K, completely bypassing the machine-building phase that every other major economy considered essential.

This wasn’t just an oversight. It was a strategic blind spot that left us “basically renting out our future to countries outside,” as one entrepreneur puts it.

..

But then something shifted in the workshops

Come 2014, with Make in India. Come the Capital Goods policy, PLI schemes, and the broader calls for Aatmanirbhar Bharat.

The missing middle started getting built. One machine at a time.

The torchbearers emerged from unexpected places:

Established names like Jyoti CNC and Ace Micromatic began expanding. Newer innovators like Ethereal Machines started breaking ground with five-axis CNC technology.

And in workshops scattered across India’s industrial map, a quiet revolution began humming to life.

..

The Haridwar story that changes everything

Picture this: A young industrial designer who never wanted to join his family’s spring-making business.

Vishakh Ranotra had other plans. Furniture startups. Custom motorcycles in Jaipur. Anything but the family factory.

Then something clicked during his six-month stint building custom bikes.

“If these things are possible, if I can make this physical product, work on the lathe machine and make all these things, why not take this back and try to re-engineer all these machines we have in our factory?”

The spark that started everything: Working 12-hour days as essentially a mechanic, building motorcycle parts by hand, Ranotra realized he could reverse-engineer the imported Chinese machines his family business relied on.

Five months ago, he made a dramatic decision. Left Delhi permanently. Moved to Haridwar with his wife.

“I wanted to be right in the middle of it, where I can step in, how we can scale this up, what opportunities we can find along the way.”

The results speak for themselves:

⭐ Four-axis spring coiling machines operating at nearly double the speed of imported two-axis alternatives

⭐ Costs 10-20% less than comparable Chinese machines

⭐ Zero benefits of scale manufacturing, yet competitive on both speed and cost

But here’s the challenge that reveals everything: Even Ranotra’s breakthrough innovations depend on imported “brains and muscles” - the precision servo motors and CNC controllers that no Indian company has successfully cracked.

..

The Bengaluru breakthrough that surprised the world

2,300 kilometers south, two young engineers were taking a completely different approach.

No reverse engineering. No family business foundation. No money to buy existing machines to study.

Kaushik Mudda and Navin Jain at Ethereal Machines had to build everything from theoretical first principles.

The mathematics nearly broke them: “It actually took us a good 18 to 24 months just to get the maths right. We had to do very complex matrix transformations. It was a very hard balance.”

When their five-axis machines were ready, they faced the credibility wall every Indian hardware startup knows too well:

“We started facing a lot of resistance. It’s impossible that an Indian company could make this, let alone two youngsters. The company almost came to the brink of death.”

Enter the pivot that changed everything: Manufacturing as a service.

Instead of selling machines, they started making components with their machines. “Now I make it with my machine and with my hands, customers are usually indifferent to it.”

Today, their 300,000 square-foot factory in Yeshwantpur runs 70 machines round-the-clock.

The breakthrough specifications that matter:

⭐ Aura three-axis system: ±4 micrometers positional accuracy (one-twentieth of human hair thickness)

⭐ Nimbus five-axis platform: mills aluminum, steel, titanium, Inconel to tight tolerances

⭐ Output delivered at one-third the cost of similar machines worldwide

⭐ 70% of revenues from exports across Europe, America, and Middle East

The irony? Their first sales came from Italy and Japan, not India.

..

The manufacturing imperative that changes everything

“For India, being a manufacturing hub is not just a choice. It’s an absolute necessity if we really want to be a dominant player in the geopolitical landscape.”

This isn’t just entrepreneur rhetoric. The numbers are backing up the narrative shift.

Manufacturing sector GVA growth jumped to 11.89% in FY 2023-24, up from 7.3% the previous year.

This isn’t incremental growth. This is acceleration.

The employment reality no one talks about: “Forty-five percent of our population is below age 29. If sustainable jobs have to be created across that huge segment, it’s going to come only from manufacturing.”

Software companies optimize for lean teams. WhatsApp had 20 employees when acquired. Manufacturing, by contrast, creates wide-based employment by design.

The cultural shift that’s finally happening:

⭐ Over 3,600 “deep tech” startups in India, including hardware, IoT, robotics

⭐ Public markets treating companies like Azad Engineering, Unimech Aerospace, BDL, HAL favorably

⭐ FDI increasing in manufacturing, especially electronics and automotive

⭐ Government pushing to double manufacturing’s GDP share from 12% to 23% over two decades

“The way they are pushing it is phenomenal. I’ve never seen this kind of push in the history of this country.”

..

The bottleneck that reveals the next opportunity

But here’s what both entrepreneurs highlight as the critical missing piece:

CNC controllers. The software and electronics that serve as the “brains” of sophisticated machines.

“Motors are still available. There are good guys in India who manufacture them. But the CNC controllers, only a few people worldwide have cracked it.”

This dependency on imported intelligence represents both vulnerability and enormous opportunity.

Ethereal Machines is taking a shot at it: “Hopefully in the next 12 months, we should have India’s first CNC controller.”

..

What this means for India’s industrial future

The convergence is already visible.

AI meeting robotics meeting additive manufacturing. Predictive maintenance algorithms. Intelligent toolpath optimization. 3D metal printing capabilities.

The opportunity ahead isn’t just about catching up. It’s about leapfrogging from “build-to-print” to “design-and-build” capabilities.

While the world focused on India’s software prowess, a quiet industrial revolution began in workshops and factories across the country. Machine makers who could reshape everything about how India manufactures, exports, and competes globally.

The question isn’t whether India can build world-class machines anymore.

The question is how fast this silent revolution scales.

And whether the next generation recognizes that sometimes the most powerful transformations happen not in Silicon Valley offices, but in the workshops where sparks fly and precision matters down to the micrometer.

The machine makers are here. The revolution is already humming.

Thank you for reading Sharp by Swarajya. You can read Karan Kamble’s original report here.

The Strategic Reality AI foundational models aren't neutral - they embed the values, languages, and priorities of their creators. The US, China, and EU understand this: whoever controls foundational models controls global discourse and decision-making.

• India, despite 1.4 billion people and world-class tech talent, remains on the periphery because it lacks the fundamental infrastructure to build sovereign AI.

The Compute Bottleneck: Numbers That Matter Training a 600-billion parameter multimodal model requires:

• 20,000-30,000 NVIDIA H100 GPUs running for 60 days

• 36 sextillion FLOPS of computation

• Example: xAI's Grok 3 used 100,000 H100 GPUs, consuming 200 million GPU-hours

As AI evolves toward physical robotics and autonomous systems, these demands will only intensify. OpenAI's "Stargate" initiative plans continental-scale supercomputers for artificial general intelligence.

The Semiconductor Stranglehold The US operates a three-tier export system:

• Tier 1 (UK, France, EU): Minimal restrictions on high-end AI chips

• Tier 2 (India): Highly regulated, ~100,000 AI chip annual cap

• Tier 3 (Russia, China): Severe restrictions or outright denial

India faces potential relegation to Tier 3 as geopolitical dynamics shift. NVIDIA's dominance compounds this vulnerability - they maintain massive backlogs with US companies getting priority, leaving countries like India as secondary customers.

China's $130 Billion Lesson China anticipated US restrictions and built comprehensive domestic capabilities:

Industrial Infrastructure:

• Manufacturing equipment: SMEE, AMEC, Naura Technologies

• Design tools: Empyrean, Xpeedic, Primarius (competing with Siemens/Cadence)

• Chip companies: Huawei (Ascend series), Moore Threads, Lisuan

Strategic Approach:

• "Big Fund" 3.0 targets entire supply chain: design, manufacturing, packaging, testing, equipment, materials

• Rapid feedback loops between toolmakers and strategic companies (SMIC, Huawei, YMTC)

• Innovative approaches: developing EUV mirrors through space-based solar X-ray imaging projects

Result: Chinese companies now just one generation behind NVIDIA.

India's Technical Pathway

Design Foundation:

• GPU architecture fundamentals for AI/HPC workloads

• OpenLane platform: integrates Yosys (RTL synthesis), OpenROAD (placement/routing), Magic/Netgen (verification)

• Computational photolithography tools: LithographySimulator, Optolithium, OpenILT for predicting/optimizing chip patterns

Fabrication Capabilities:

• EUV Light Sources: BARC's Accelerator Division (Dr. Amitabha Roy) has soft X-ray expertise; EUV falls in 10-100nm soft X-ray band

• Laser Systems: DRDO fiber laser technology for heating tin droplets

• Precision Mirrors: BARC's Atomic & Molecular Physics Division has multilayer mirror experience for X-ray applications

• Positioning Stages: Magnetic levitation technology for nanometer-level wafer placement (research by Dr. P. Ananthababu, Dr. K.S. Vikrant)

Inspection Systems:

• Electron microscopy for defect detection (open-source SEM/TEM projects available)

• Sputtering and vapor deposition tools for manufacturing and mirror production

The Critical Gap in Current Policy India's Design Linked Incentive (DLI) scheme has fundamental flaws:

• Lacks clearly stated technical goals

• Focuses on chip fabrication while neglecting machine tools for design/manufacturing

• Produces laboratory projects without commercial viability

• Insufficient technical depth in problem statements

A National Mission Framework

Talent Mobilization:

• Recruit Indian diaspora from global semiconductor firms as mentors/leaders

• Target outreach programs offering opportunities to contribute to national sovereignty

• University-level students integral to mission for hands-on skilled workforce

Industrial Integration:

• Agrani Labs (GPU development), InCore Semiconductors, Tata-PSMC fab as first customers

• Industrial feedback essential for complete development and commercial deployment

• Seamless collaboration between diaspora, academia, engineers, startups, MSMEs

Strategic Problem Statements:

• Computational photolithography simulators with inverse lithography and ML acceleration

• Universal chip architecture with ML algorithm software stacks

• Language model-based generative design for photomask generation

• End-to-end integrated chip design/verification tools with ML workflows

The Geopolitical Window Current conditions create unprecedented opportunity:

• Technological democratization through open-source tools

• Substantial talent pool and global diaspora

• Growing industry demand and geopolitical tensions highlighting supply chain vulnerabilities

Bottom Line India faces an existential choice: continue dependence on increasingly restricted foreign supply chains or achieve technological self-reliance through indigenous semiconductor capabilities.

• The compute bottleneck threatening AI sovereignty can only be resolved by building the full semiconductor stack - from design tools to fabrication equipment to inspection systems.

• The tools exist, the talent is available, and the geopolitical moment is right.

• What's missing is comprehensive national will and coordinated execution.

A fuller version of this draft is available on our website at this page.

Arpan Sow holds a Master of Science degree in Aerospace Engineering from Technion, Israel, and is currently a PhD candidate there. He is an avid aerospace propulsion and semiconductor enthusiast.