FADU Inc.

Nearly all revenue comes from selling physical units at the point of delivery rather than being earned gradually over a service period, so it behaves as a unit-sale business rather than a subscription-like one. Within that, the controller chip itself is the larger source of revenue, with the finished drives built around it contributing a smaller share.

For a chip designer that builds nothing itself, growth comes from winning design slots with customers and carrying each new chip generation through their qualification process into mass production, rather than from adding factories or headcount. Each generation requires heavy research spending well before it is known whether it will convert into revenue, and on the evidence available, total spending has consistently run ahead of revenue rather than the reverse.

For every physical product it sells, it depends entirely on outside partners: an outside design house supplies the controller chips and separate contract manufacturers assemble the finished drives, since the company owns no production facilities of its own. Most electronic components come from multiple suppliers, but its two most critical inputs, NAND flash memory and DRAM, are bought through a concentrated, oligopolistic market dominated by a small number of large global suppliers, and CompanyGraph separately maps it as sitting downstream of a wide band of other industries that feed into it.

Its buyers sit within the storage supply chain itself: memory manufacturers, other SSD makers, and large technology companies, whose end products run in hyperscale data centers and enterprise servers. Demand within that buyer base is uneven. Its own disclosures show a single, undisclosed customer accounts for most of its revenue, and it names one supply counterparty, Macnica Galaxy Inc., within its contracted sales relationships. CompanyGraph separately maps it as supplying a narrower band of downstream industries than the number of industries that feed into it.

Structurally, the way it operates, designing a component and coordinating outside partners to manufacture and qualify it, is a common shape shared by a large number of similar companies, not a distinctive one. The company separately claims specific technical performance advantages over rivals in its own materials, but those are the company's own claims rather than something independently verified, and they say nothing about whether competitors could replicate them. Structurally near is not the same as moving together or being interchangeable, it means CompanyGraph sees a shared way of operating or a detected pattern, not a price relationship or a comparison verdict.

Before a customer buys, it must first put a candidate chip through the customer's own technical qualification process, checking it against engineering requirements, and only after that process is complete does supply move into negotiated schedules and mass production. That qualification effort is a sunk cost for both sides once complete, and the company's own disclosures show a multi-year book of already-committed orders extending well into the future, evidence that at least some customers have moved past qualification into a standing supply relationship.

The usual limit for a chip producer, a fixed physical plant that caps how much it can make, does not apply directly here: the company's own disclosures show it owns no production facilities and instead buys manufacturing from outside partners, so any physical capacity ceiling sits with those partners rather than with the company itself. What its own account does point to as limiting is sequence and dependence: each new chip generation requires a large research investment before it can earn anything, revenue only begins after a customer's own qualification process is completed, and the resulting revenue leans heavily on a small number of large customers and on conditions in the broader memory-chip market.

The company's own disclosures point to concentration as a named risk: a single, undisclosed customer accounts for most of its revenue, so a change in that one relationship would affect an outsized share of the business rather than being absorbed across many buyers. Its own filings also tie product demand directly to conditions in the broader memory-semiconductor market, a cyclical market it does not control, so the effect of any change in that one customer relationship sits on top of a wider industry cycle it cannot set the terms of.

It operates under securities-market regulation and stock-exchange listing rules in its home market, and its own disclosures report pending securities-related legal claims and a criminal case against the company and certain of its executives over alleged violations of capital-markets law. It also tracks currency movements, principally in the US dollar, as a financial risk, and states that demand for its products cannot be separated from broader conditions in the memory-chip market, a market it does not set the terms of.