The headline was almost a hundred-percent growth. The operating story is something else entirely. Ambiq Micro's second quarter delivered net sales of $33.9 million, up 89.7% from a year-ago $17.9 million and the company's first $30 million-plus quarter as a public company. The first-half total of $59.0 million was already 81% of full-year 2025 revenue of $72.5 million, a step-function lift that management attributes to new product launch ramps and a "new major customer product launch which led to significant sales in both" the quarter and the half. Gross profit more than doubled to $15.3 million and the gross margin expanded 4.9 percentage points to 45.0%, the highest in the company's reported history. Those are real numbers, not a mirage. The loss line, by contrast, tells a different story. Net loss was $7.1 million - narrower than the year-ago $8.5 million on an absolute basis, but the loss-from-operations print was almost flat at $(8.7) million versus $(8.8) million a year ago, even as revenue grew 90%. The difference between the two is interest income earned on a cash pile that was reloaded twice in six months. Strip the interest, and the company is still spending roughly the same on operations it spent a year ago on a revenue base less than half the size. The growth is real; the operating leverage is not yet visible.
The setup behind the headline is a fabless semiconductor designer that went public in late July 2025, sold a follow-on in January 2026 at $31 a share, and sold another follow-on in June 2026 at $78 - each time roughly when its own cash runway started to look short against an operating loss still in the high-single-digit millions a quarter. Cash and equivalents now sit at $366.8 million, against an accumulated deficit of $374.0 million and a stockholders' equity of $404.7 million. The company carries no long-term debt; its liabilities are essentially trade payables, accruals, and operating-lease commitments. At a reference price of $65.96, market capitalization is approximately $1.59 billion, against trailing revenue on a first-half-annualized basis of roughly $118 million, a P/S near 13.5x - rich in absolute terms and rich relative to other recent-issue semiconductor designers, but defensible against a backlog and pipeline the company is unwilling to quantify in dollars. The thesis is a bet on three things: that the gross-margin step-up is structural, that revenue growth can keep running above 50% into 2027, and that customer concentration - top three end customers at 78% of Q2 revenue and 75% of H1 - does not become a single-customer cliff. None of those has a proof date attached.