A-Portfolio

The company in one breath

Ambiq Micro designs tiny computer chips for watches, medical gear, smart glasses, and factory sensors. Device makers pay Ambiq because its chips can run artificial intelligence inside the device without chewing through the battery like a teenager through a refrigerator.

Ambiq does not own a chip factory. It designs the brains, then Taiwan Semiconductor Manufacturing Company, or TSMC, makes every wafer. That keeps factory costs off Ambiq's books, but it also gives one supplier control over every chip Ambiq can sell.

Why now

Ambiq went public at $24 a share in July 2025. That means the market has only had about a year to judge this business, and a good recent listing should use that first year to prove the pitch was real.

The second quarter supplied proof. Sales reached $33.9 million, up 89.7 percent from a year earlier. That means customers nearly doubled what they bought, and good for a young chip designer would be growth above 30 percent without one customer doing all the lifting.

Management expects third-quarter sales of $36 million to $37 million. That means another sequential increase, which is trader language for beating the immediately preceding quarter, and good would be landing inside that range while shipping enough product to meet demand.

The stock was $65.70 at 3:09:43 p.m. Eastern today on the Robinhood broker feed. That means buyers pay 173.8 percent more than the IPO crowd paid, and good entry quality would look like either a lower price or much more proven profit.

The numbers, translated

Revenue for the first half reached $59.0 million, up 75.5 percent. Revenue is the money customers send in before costs, so this says the growth was bigger than one lucky quarter. Good would be several more quarters above 30 percent while the customer list widens.

Gross margin reached 45.0 percent, up from 40.1 percent. Gross margin is what Ambiq keeps after the direct cost of making its chips, so it kept 45 cents from each sales dollar before paying engineers and office bills. Good for a specialized chip designer would be 50 cents or more as newer products mature.

The operating loss was $8.7 million, barely better than $8.8 million a year ago. Operating loss is what the core business lost after normal expenses, so nearly doubled sales did not yet produce a much smaller dollar loss. Good would be a loss below 15 cents for each sales dollar, then an actual profit.

Ambiq used $20.7 million of cash in operations during the first half, almost double the $10.5 million used a year earlier. That means inventory and the daily business consumed more cash even as sales grew. Good would be cash use falling toward zero because customers, not shareholders, fund expansion.

Cash stood at $366.8 million against $31.0 million of total liabilities. That means Ambiq has a huge financial cushion and little owed to outsiders. Good is exactly this kind of runway, but most of the cushion came from selling more stock, not from business profit.

At today's $65.70 price, I calculate an operating-business value of about $1.22 billion, or roughly 9.0 times the current annualized sales pace. Operating-business value, also called enterprise value, strips out the cash pile, and annualized sales simply multiply the latest quarter by four. Good entry math for this unprofitable company would be closer to six times sales unless growth and profit improve fast.

What I like

I like the technical wedge. Ambiq says its SPOT design can use two to five times less power than conventional chip designs. That means a device can think locally for longer without a recharge, and good would be more medical and industrial customers proving the advantage matters beyond watches.

I like that gross profit grew 112.9 percent while sales grew 89.7 percent. That means Ambiq kept more money from each new sales dollar instead of buying growth with discounts. Good would be gross profit continuing to outrun sales as product prices and factory efficiency improve.

I like the balance sheet. $366.8 million of cash gives management years to hire engineers, secure supply, and build software. Good would be using that cushion once, then letting the business carry itself instead of returning to shareholders with another empty glass.

What worries me

Three end customers supplied 28.9 percent, 24.4 percent, and 24.2 percent of second-quarter sales, or 77.5 percent combined. That means almost four of every five sales dollars depended on three buyers. Good would be the top three falling below 60 percent because new customers grow, not because an old one leaves.

TSMC makes 100 percent of Ambiq's wafers. That means one manufacturing partner controls all supply. Good would be a qualified backup source or enough reserved capacity that a squeeze cannot erase customer demand.

Shares outstanding rose to 24.15 million from 18.32 million at year-end, about 31.8 percent more slices of the company. That means Ambiq raised cash by making every existing slice smaller. Good would be sales and gross profit per share growing faster than the share count, with no third stock sale this year.

The social proof is basically a ghost town. The last-30-days scan found just two tiny YouTube videos totaling 18 views and no useful Ambiq discussion on Reddit, Hacker News, GitHub, or Polymarket. That means the filing carries the case, not crowd enthusiasm, and good would be credible technical users discussing real deployments rather than stock-promotion confetti.

What would make this a buy

My price trigger is $50 to $55 a share. At that range, the operating business would cost roughly 6.2 to 7.1 times the current annualized sales pace, which gives the growth story more room to beat the price. Good would be reaching that range without a broken customer win or supply problem.

My proof trigger is two more quarters with sales growth above 50 percent, gross margin at or above 45 percent, operating loss below 15 percent of sales, and the top three customers below 70 percent of sales. Those four tests mean growth stayed fast, each sale stayed valuable, losses started shrinking, and the company became less fragile. Good is all four together. Passing one test while failing three is not a buy case, it is a magic trick.

What makes me walk away

I walk if sales growth drops below 30 percent for two straight quarters before the company reaches operating profit. That means the growth engine slowed before the business learned to make money. Good would be staying above 30 percent until profits arrive.

I walk if gross margin drops below 40 percent for two quarters. That means Ambiq is losing pricing power or paying too much to make chips. Good would be holding 45 percent now and working toward 50 percent.

I walk if the top three customers still supply more than 80 percent of sales by mid-2027. That means management failed to broaden the business after raising enough cash to do it. Good would be less than 60 percent with new medical, industrial, and smart-device programs filling the gap.

I walk if the share count grows more than 15 percent in another year without faster growth in sales and gross profit per share. That means shareholders keep paying for expansion while their ownership shrinks. Good would be little or no dilution until the core business funds itself.

Earnings On Deck

AMBQ, August 11: Ambiq already reported this week. Watch third-quarter sales guidance of $36 million to $37 million, which means management expects a sixth straight quarter of growth, and good would be landing in that range with gross margin still above 45 percent. A bad print next quarter would miss the low end by more than 10 percent or show that supply problems, not demand, control the business.

AMAT, August 13 after the close: Applied Materials sells the equipment chip factories use to build semiconductors. Watch demand from leading-edge factories, which means plants making the newest chips, and good would be stable orders plus confident guidance despite export limits. A bad print would show customers delaying equipment purchases, because that can warn that the chip boom is losing fuel upstream.

The verdict

73 out of 100. Watch. Ambiq earned points for a real low-power advantage, 89.7 percent sales growth, a 45.0 percent gross margin, and $366.8 million of cash. It lost points because three customers control 77.5 percent of sales, TSMC controls all production, the business still burns cash, shareholders just ate 31.8 percent dilution, and the stock costs about 9.0 times annualized sales.

A score of 73 puts it in the watch bucket. Candidate status starts at 80, which means the evidence and entry price must both clear a higher bar. I want the company. I do not want today's bill.

The DBTech read

Enterprise value is the price of the operating business after subtracting cash and adding debt. Ambiq's stock market value is about $1.59 billion, which means that is the price attached to all common shares, while subtracting its cash leaves about $1.22 billion for the business itself. Good analysis uses enterprise value when two companies carry very different cash piles. Think of buying a bar for $1 million and finding $200,000 in its safe, the bar really cost you $800,000. Cash changes the bill, but it does not make a weak business strong.

Sources and as-of date

As of August 13, 2026. AMBQ last traded at $65.70 at 3:09:43 p.m. Eastern, from the Robinhood broker feed. That means the price was live within minutes of writing, and good source discipline means every market price carries both a clock and a named feed. The official August 12 close was $65.59, from the SIP listing-exchange close supplied by Robinhood.

1. Ambiq Q2 2026 Form 10-Q, filed August 11, 2026: https://www.sec.gov/Archives/edgar/data/1500412/000119312526344145/ambq-20260630.htm

2. Ambiq Q2 2026 earnings release, August 11, 2026: https://www.sec.gov/Archives/edgar/data/1500412/000119312526343443/ambq-ex99_1.htm

3. Ambiq official site and product description: https://ambiq.com/

4. Robinhood broker feed and official prior-session close: https://robinhood.com/us/en/stocks/AMBQ/

5. Last-30-days community scan, run August 13, 2026: /home/derek/.hermes/profiles/bobby/workspace/newsletters/a-portfolio/research/ambiq-micro-ambq-stock-earnings-raw-rewrite.md

6. Nasdaq August 2026 IPO calendar, checked August 13, 2026: https://api.nasdaq.com/api/ipo/calendar?date=2026-08