India's 200 Million Chip Mirage: The Low-End OSAT That Won't Secure Any Supply Chain
The press release landed with the weight of a national milestone. CG Power, an Indian electrical equipment manufacturer, announced it had commenced semiconductor production. The headline number: 200 million chips per year. The narrative: a step toward supply chain resilience. The reality, after a forensic teardown of available data and industry patterns, is far less revolutionary. This is not a fab. This is not advanced manufacturing. This is a government-subsidized OSAT line—an assembly and test facility—that adds marginal value to a global chip ecosystem already drowning in low-end capacity.
Proof exists; it is merely waiting to be verified. The verification begins with the number itself. Two hundred million units annually translates to roughly 550,000 chips per day. For a foundry, that would be a fraction of a single wafer run. TSMC’s Fab 18 produces over 100,000 12-inch wafers per month; each wafer yields thousands of chips. A foundry operation with a 200 million annual output would be operating at less than 1% of a mainstream logic fab. The math does not support a wafer fabrication narrative. Even for a mid-sized OSAT, 200 million units is modest—ASE Technology processes over 10 billion chips annually. The scale here is at the very bottom of the industry pyramid.
The technology node remains unstated, which is itself a signal. No nanometer, no wafer size, no mention of lithography or etching equipment. In foundry announcements, process node is the first data point disclosed. Its absence here indicates that no meaningful transistor fabrication is occurring. The most likely scenario is a package assembly line: imported die are bonded onto lead frames, encapsulated in plastic, tested, and shipped. This is the semiconductor equivalent of a car manufacturer that only installs seats and tires—the engine, transmission, and chassis are built elsewhere. The Indian government’s subsidy scheme covers 50% of capital expenditure for OSAT facilities; CG Power is leveraging this to diversify from its core power equipment business. The chips being produced are almost certainly discrete power devices or simple integrated circuits for industrial use—IGBT modules, MOSFETs, or LEDs. These are mature, commoditized products with low margins and high competition.
Contrast with the global competitive landscape. The top 10 OSAT players—ASE, Amkor, JCET, Powertech, etc.—control over 80% of the market. Their average gross margins hover around 18-22% for standard packaging, and that is with decades of yield optimization and scale. CG Power enters with zero prior semiconductor experience, no certified automotive qualification (IATF 16949 is a multi-year journey), and a home market that still imports over 90% of its chip demand. The only edge is government capital: the Semiconductor India Program promises 50% project cost support. Without that, the business case collapses. A typical low-end OSAT line requires $100-$200 million investment; annual depreciation alone would eat through any profit on 200 million low-cost devices. Break-even capacity utilization is likely above 70%, yet initial utilization for new entrants often starts below 40%. Red ink is baked into the first two years of operations.
Supply chain resilience is the stated goal, but the analysis reveals a contradiction. The facility relies on imported die, lead frames, bonding wire, and molding compounds from Japan, Taiwan, and China. The “Made in India” label applies only to the last 15% of value addition—assembly and test. Global semiconductor supply chains are vulnerable at the wafer fabrication stage; OSAT capacity is abundant in Southeast Asia (Malaysia, Thailand, Vietnam) and China. Adding a small OSAT in India does nothing to insulate against a shortage of advanced chips or a disruption in raw wafer supply. It merely shifts the final packaging location from one non-critical node to another. The claim of enhancing global resilience is a species of marketing speak—a lead frame in India is as easy to replace as a lead frame in Penang.
The algorithm remembers what the witness forgets. The witness here is the Indian government’s press machinery, eager to tout any semiconductor announcement as proof of strategic progress. But the algorithm—the cold logic of capacity math, cost structures, and technology roadmaps—remembers a history of failed attempts. India has announced over a dozen semiconductor projects in the past two decades: Fab City in Hyderabad, Hindustan Semiconductor in Bengaluru, HSMC with STMicroelectronics, and multiple others. None reached volume production. The pattern is consistent: political urgency meets industrial reality, and reality wins. CG Power’s announcement follows the same script—a low-barrier entry point (OSAT), heavy government subsidy, and a timeline that looks aggressive on paper but is easily delayed. The difference this time is geopolitics: the US and its allies want to reduce dependence on China and Taiwan, even for low-end packaging. India is a candidate, but so are Malaysia and Vietnam, which already have skilled workforces, established supply chains, and lower labor costs. The competitive advantage is slim.
Bulls will point to India’s growing domestic demand for power semiconductors, driven by renewable energy, electric vehicles, and industrial automation. That is a genuine tailwind. The Indian power module market is projected to reach $3 billion by 2030. CG Power, with its existing customer base in the power equipment sector, could capture a share of that demand. If the company secures a few large anchor clients—say, Tata Motors for EV inverters or Adani Green for solar inverters—the line could run at decent utilization. Government contracts for defense or smart meters could provide a baseline. In this bullish scenario, the OSAT becomes a viable niche player, generating modest but stable cash flows. The contrarian view is not that the project will fail entirely, but that its impact will remain marginal. A 200 million chip facility, even at full capacity, represents less than 0.1% of global OSAT output. It does not move the needle on supply chain resilience, does not reduce India’s chip import bill by more than a rounding error, and does not signal technological sovereignty. It is a small step, dressed as a giant leap.
Ledgers balance, but ethics remain uncalculated. The true cost of this project is not the capital expenditure but the opportunity cost. India’s semiconductor policy allocated $10 billion in incentives. This OSAT line consumes a portion of that fund, funds that could have been directed toward building a viable fab ecosystem—training engineers, constructing cleanrooms, subsidizing EDA licenses—or toward more impactful sectors like electronics manufacturing. Instead, the money underwrites a project that, at best, will serve as a testing ground for future, more ambitious endeavors. At worst, it will become another white elephant, sustained by subsidies until the political will evaporates. The cynical interpretation is that CG Power is a convenient vehicle for the government to show progress, and the company is using the subsidy to build an asset that can be sold or repurposed later. The math supports the cynicism: without the subsidy, the internal rate of return on a $150 million OSAT at 20% gross margin and 60% utilization is below 5%, far below India’s cost of capital. No rational private investor would proceed.
The takeaway is a question, not a conclusion. How many such announcements will it take before the market learns to read the technical details before the press release? The next time a headline screams “India enters chip manufacturing,” check for wafer size, for process node, for committed customer orders, for the absence of subsidy dependency. Until those numbers are public, assume the announcement is a political artifact, not an industrial breakthrough. The algorithm never lies. The PR department does.