Ideal Power is a fabless semiconductor designer whose entire equity is an option on whether a patented bidirectional power switch named B-TRAN graduates from evaluation kits into volume circuit-protection hardware. What changed in the June quarter is not the income statement. What changed is the company's ability to stay solvent long enough for customers to finish testing, after two equity raises in the first half recapitalized the balance sheet and a foundry in Asia produced first functional silicon. The market still prices that option as if conversion remains unproven, which is the honest reading of a quarter that printed almost no product revenue.
The most important recent development is the long-term wafer supply agreement with an automotive-qualified foundry outside China, paired with the second registered-direct raise that closed in May. The mechanism is straightforward. Without a high-volume fab that can hit cost targets, even a winning design stays a science project. Without cash, the company cannot fund the sales and applications engineers who move customers through qualification. First silicon at that foundry is the first physical proof that the process is transferable off a development line. The May financing, led by existing institutions, bought years of runway at the current burn rate and simultaneously diluted the share count enough that any later production win is spread across a much larger base.
The tension is that the commercial story is still almost entirely qualitative. Revenue in the June quarter was $5800. That figure is a rounding error against a cash pile of $41 million and against a sales funnel that management now describes as more than $400 million of opportunity. A growing funnel is not a backlog. The Asia solid-state circuit breaker customer, the Stellantis electric-vehicle contactor program, and the hyperscaler prototype tied to the NVIDIA Rubin Ultra architecture are all still in evaluation. The strongest argument against the equity is that Ideal Power has spent years announcing design activity while the income statement has never shown a commercial ramp, and the latest recapitalization simply extended that clock.
The next several months resolve whether any of those named programs produce a production order. Low-volume shipments to support the Asia customer's prototype builds are the first observable test, along with delivery of the hyperscaler-oriented breaker prototype by year-end. If those remain samples, the equity stays a cash-backed option. If a production purchase order appears, the market has to reprice a company that today is valued at little more than its cash plus a thin technology premium.