$AAOI: The Laser Maker Running Faster Than Its Own Margins

Five consecutive quarters of record revenue. An 800G product line that more than doubled sequentially in Q2 2026. A single hyperscale order worth more than $200 million sitting on the books. By almost any measure of momentum, Applied Optoelectronics looks like the kind of company that should be the easiest pitch in the room.

Then you open the balance sheet. Three equity raises in one year. A share count that has expanded roughly 44% year over year. A gross margin stubbornly pinned at 29.8% while management talks about a 40% long term target. The stock fell more than 30% in the weeks after the company announced its third $600 million ATM program in August, before investors had even seen a single shipment from the 1.6T product line.

We think the real question here is not whether $AAOI is capturing the AI infrastructure wave. It clearly is. The question is whether the per share economics of that capture are survivable as the company funds an unprecedented capacity build by continuously printing new stock. That tension, between a genuine demand story and a dilution machine running in parallel, is what makes this one worth examining carefully.

The 60 second version

• Applied Optoelectronics ($AAOI) designs and manufactures optical transceivers, lasers, and related components, selling primarily to hyperscale data centers and cable operators from its Texas and Taiwan facilities.
• The shift: AI clusters need vastly more optical bandwidth per rack, pushing the industry from 400G to 800G and then 1.6T transceivers, a generational product cycle that rewards suppliers who can actually manufacture at scale.
• The number: Q2 2026 revenue of $191.9 million, up 86% year over year, with management guiding Q4 2026 revenue to exceed $500 million, implying an annualized run rate above $2 billion entering 2027.
• The catalyst: Initial 1.6T transceiver shipments are guided for the end of Q3 2026; the Q3 earnings report, expected November 5, 2026, is the first moment investors can verify whether that ramp is real and whether the $255, $290 million Q3 revenue guide was hit.
• The main risk: Serial equity dilution. Three large scale financing rounds in 2026, including a third $600 million ATM announced August 21, with no clear ceiling on the share count expansion and gross margins still below 30%.
• Market capitalization: approximately $8.96 billion as of September 9, 2026, which fails our sub-$5B screen, disclosed above.

Here is the sharpest fact in the file: management guided Q4 2026 revenue above $500 million. If that number lands, the company will have grown from roughly $100 million per quarter to more than $500 million per quarter in six quarters. That is not a typical technology ramp. That is a near vertical demand curve meeting a supplier who built the factory before everyone else showed up.

The counterintuitive piece is the Amazon warrant. Buried in the company’s filings is a warrant for Amazon to acquire up to 7.9 million shares of $AAOI stock. That is an anchor relationship, not a spot purchase. Hyperscalers do not hand out equity warrants to transceivers suppliers they plan to replace next cycle.

Our thesis is this: $AAOI holds a genuine manufacturing and technology lead in 800G and 1.6T optical transceivers at exactly the moment AI infrastructure spending is accelerating, but the per share value of that lead is being eroded in real time by an equity issuance pace that has expanded the share count by roughly 44% in the past year alone, with a third $600 million ATM program still open as of this writing.

The stock is worth understanding in detail. Whether it is worth owning at the current price and share count trajectory is a harder call, and we will show you exactly where we land.

The shift

The constraint inside AI data centers for the past three years has not been compute. It has been bandwidth, specifically the optical bandwidth connecting GPU clusters to each other and to the network. When a hyperscaler trains a large model across thousands of accelerators, every chip is constantly sending data to every other chip. The speed at which that data moves, and the distance it can travel without degradation, determines how efficiently the cluster runs.

The industry answer has been to push transceiver speeds upward in roughly two year cycles: from 100G to 400G, then to 800G, now to 1.6T. Each jump is not just faster, it requires meaningfully different laser technology, different packaging, and in the case of co packaged optics, a different physical relationship between the optic and the switch silicon. Suppliers who can only build the previous generation get commoditized quickly. Suppliers who can ship the next generation at volume before demand peaks capture the margin.

The binding constraint that eased was manufacturing scale in the United States. For most of the past decade, optical transceiver manufacturing was concentrated in Asia, particularly in China and Taiwan, at companies like HG Genuine and Innolight. The combination of AI driven demand, hyperscaler supply chain anxiety after years of pandemic and geopolitical disruption, and explicit U.S. government pressure to onshore critical technology infrastructure created an opening for a domestic supplier who had the process knowledge to scale quickly. That opening is not theoretical. It is already in the revenue numbers.

The specific inflection arrived in late 2025 and accelerated through the first half of 2026. Hyperscalers who had been ordering 400G modules in bulk shifted their planning conversations to 800G and began placing 1.6T qualification orders. The companies positioned to capture that shift are few, and the one with the largest U.S. manufacturing footprint and the deepest existing hyperscale relationships is $AAOI.

The company

Applied Optoelectronics was founded in 1997 in Sugar Land, Texas, by CEO Thompson Lin, who still runs the company and still holds meaningful equity. It started as a laser and component business serving cable television operators, which is why CATV still accounts for roughly 42% of revenue today, a segment most investors ignore because it is slow and mature. The data center segment, which is where all the growth lives, was built on the same underlying laser and packaging expertise, applied to a different market.

What the company actually does is straightforward: it takes semiconductor lasers, builds them into optical subassemblies, packages those into transceivers and modules, and sells them to whoever needs to move data over fiber at very high speeds. It manufactures in Sugar Land and in Taiwan, and it builds a significant portion of its own production machinery in house, which matters more than it sounds. When the industry hit equipment supply bottlenecks during the pandemic, $AAOI was less exposed than competitors who depended on third party tooling vendors.

The specific technology detail that stands out from the filings: in December 2025, the company introduced a 400-milliwatt narrow linewidth pump laser aimed at the silicon photonics and co packaged optics market. The gross margin on these external light source products, cited on the earnings call, is expected to run at 55, 65% as production ramps in 2027, which would be roughly double the company’s current blended margin. That product is not yet contributing meaningfully to revenue, but it is the margin lever the whole long term story hinges on.

An aside worth making: Thompson Lin’s purchasing activity over the past six months is an unusual signal. He made six separate open market purchases totaling 26,675 shares for roughly $597,470, a modest sum relative to the company’s scale but notable because he is buying while insiders collectively have sold more than $78 million of stock in the same period. The CEO buying into his own dilution is either very confident or very early. We do not know which yet.

The numbers and what the street expects

Q2 2026 revenue came in at $191.9 million, up 86% from $103.0 million in Q2 2025 and up 27% sequentially from $151.1 million in Q1 2026. That was the fifth consecutive record quarter. Data center revenue was $107.7 million, or 56.1% of the total; CATV was $80.6 million, or 42.0%. The 800G product line more than doubled sequentially within the data center segment and management said it expects 800G revenue to grow nearly five times in Q3.

Gross margins have not moved. Non GAAP gross margin was approximately 29.8% in Q2, flat from Q1 despite a 27% sequential revenue gain. That is the number that should concern anyone building a long term model. An 86% revenue increase has produced essentially no operating leverage at the gross line. Management targets 40% non GAAP gross margins over time, but the path there depends almost entirely on the CPO external light source product ramping at scale, which is a 2027 event at the earliest.

The company returned to non GAAP profitability in Q2, posting non GAAP EPS of $0.06. GAAP results remain in the red because of stock based compensation, depreciation on an accelerating capex base, and interest expense on the convertible notes. The $124.9 million of 2.75% Convertible Senior Notes due 2030 carry a conversion price of approximately $43.31, which at the current stock price creates meaningful dilution optionality if the notes convert.

Cash as of June 30, 2026 was $499.7 million, plus $9.0 million restricted. That sounds comfortable until you look at where it came from: financing activities provided $980.3 million in the first half, while operations consumed $73.8 million and investing consumed $633.7 million. The company spent $565.5 million on capital investments in Q2 alone. This is not a business funding its growth from operations. It is building factories on equity capital and hoping the margin improves before the market loses patience.

Management’s Q3 2026 guidance is revenue of $255, $290 million, non GAAP gross margin of 29, 30.5%, and non GAAP net income of $10.1, $24.0 million, implying a non GAAP EPS range of roughly $0.11, $0.26. The midpoint of the revenue range is $272.5 million, which aligns almost exactly with the Wall Street consensus of approximately $271 million from 13 covering analysts. On earnings, however, management’s midpoint of around $0.19 sits above the consensus of $0.14, a gap worth watching.

The full year 2026 picture is where the real divergence lives. Management is guiding revenue above $1.1 billion with operating income above $140 million. The current street consensus for fiscal 2026 sits closer to $962, $988 million, putting management above consensus by roughly 11, 14%. That is a meaningful gap, and it points in one direction: either management knows something about Q4 demand that the street has not yet modeled, or the Q4 guide of more than $500 million in revenue will be revised down. The Q3 print on November 5 will narrow that gap substantially.

Revenue estimates were revised down 4.4% over the past 90 days, and EPS estimates were cut more sharply than revenue, suggesting analysts believe the revenue ramp is real but margin delivery is not keeping pace. Thirteen analysts cover the name; the most optimistic carries a $220 price target and the most cautious sits at $109. Short interest is 13.9% of the float, elevated but not extreme for a name this volatile, with a beta of 2.98. A two times daily leveraged ETF tied to the stock has been launched, which adds to daily volatility in ways that have nothing to do with fundamentals.

Why it wins

The durable advantage here is manufacturing depth, not product design alone. $AAOI builds a significant portion of its own production machinery in house. When rivals are waiting six months for equipment from a third party vendor, $AAOI is retooling its own lines. That matters enormously in a cycle where the product generation changes every eighteen to twenty four months and the supplier who can deliver at volume first captures the contract.

The geopolitical dimension is real and underappreciated. Hyperscalers are actively trying to reduce their dependence on Chinese manufactured optical components. $AAOI is the largest U.S.-based manufacturer of AI optical transceivers. That label is worth money in procurement conversations that would otherwise be purely technical. The Amazon warrant, which entitles Amazon to purchase up to 7.9 million shares, is the clearest signal that at least one major hyperscaler views this as a strategic relationship rather than a commodity purchase.

The coming margin inflection, if it arrives, will come from the CPO external light source product. Narrow linewidth pump lasers for co packaged optics carry expected gross margins of 55, 65%, according to management commentary on the earnings call, versus the current blended rate of 29.8%. A business that is today valued on 30% margins but transitions to a product mix carrying 50% plus margins over two years would look very different to the street. The problem is that timeline remains 2027 at the earliest and depends on customer qualification cycles that are outside $AAOI’s control.

What could go wrong

The strongest bear case is not about demand. It is about the share count. In 2024 shares outstanding grew 30%. In 2025 they grew 45%. Through the first half of 2026, diluted shares reached 81.6 million, up 43.68% from the same period last year. A third $600 million ATM program was announced on August 21, 2026, and the amount actually drawn from that facility is not yet disclosed in any public filing. If the company sells the full $600 million at current prices, the share count expands further before any of the promised margin improvement arrives.

Gross margin is not improving. The company grew revenue 86% year over year and gross margin did not move a single point. Management’s 40% gross margin target requires a product mix shift that has not yet started. If the CPO external light source product is delayed, or if hyperscaler qualification trials extend longer than expected, the company could spend two more years posting 29% gross margins on a rising revenue base while continuously issuing new shares to fund the factory. In that scenario, GAAP earnings per share could actually decline even as revenue doubles again.

There is a customer concentration risk buried in the data center segment that the filings do not fully quantify. If a single hyperscaler, and the Amazon warrant implies Amazon is the largest, reduces its ordering pace, the quarterly revenue trajectory breaks. The $200 million 1.6T transceiver order from March 2026 is the most visible anchor in the pipeline, but the street does not know whether additional customers at comparable scale are in qualification or still years away from volume orders.

Finally: the stock itself. Beta of 2.98. Down more than 30% after the August ATM announcement. A leveraged ETF amplifying daily moves. This is not a name where being right about the thesis on a two year horizon protects you from a 40% drawdown in a month if the Q3 print disappoints or if the share count expansion spooks institutional holders into selling. The technical picture, a stock that has fallen sharply from its 2026 highs on dilution news and has not yet stabilized, is not a clean entry. We think the thesis is sound but the setup is messy.

The investment thesis

$AAOI holds a genuine, defensible lead in U.S.-manufactured 800G and 1.6T optical transceivers at the exact moment AI infrastructure spending is driving demand for those products at a scale the industry has never seen, but the per share economics of that lead are being actively destroyed by a share issuance program that has no disclosed ceiling and is running far ahead of any margin improvement.

What has to be true for this to work as an investment over the next twelve to eighteen months: the 1.6T product line ships on schedule, generates meaningful revenue in Q4 2026, and the $500 million plus Q4 revenue guide lands. The gross margin must start moving toward 32, 33% in the first half of 2027 as CPO product mix improves. The company must not need to draw the entire $600 million ATM before it generates enough free cash flow to fund its own capex. And the street must begin modeling the 2027 revenue run rate above $2 billion, which management has implied but which no analyst consensus yet reflects.

What the market appears to believe instead, based on the 4.4% downward revision to revenue estimates and the steeper cut to EPS estimates, is that margin delivery is impaired and the dilution risk is not priced into the growth multiple. The segment being valued at close to zero in current estimates is the CPO external light source business, which carries 55, 65% expected gross margins and does not appear in any forward model we have seen. If that product qualifies with even two hyperscale customers in 2027, the margin profile of the whole company changes materially.

The specific gap between guidance and consensus is the clearest version of the trade: management says full year 2026 revenue exceeds $1.1 billion, the street says $962, $988 million. That is roughly $120 million of revenue the street has not given them credit for. The Q3 print on November 5 and the Q4 setup described in that report will either close that gap or confirm the street was right to be skeptical.

In practice, this is a name to watch through the November 5 catalyst, not to own through it blindly. The asymmetry of the trade depends heavily on the Q3 print and, more importantly, on any commentary about Q4 1.6T volumes and whether additional hyperscalers beyond the anchor customer are beginning to qualify the product. Watch from the sideline and decide after you see whether management delivered on the $255, $290 million Q3 guide and whether the 800G five times sequential growth claim holds up in the numbers.

The thesis is wrong if Q3 gross margin comes in below 29% while the share count has continued to expand, because that combination would mean the company is spending dilutive capital to fund revenue growth that is not improving its unit economics. That single observable would tell us the margin improvement story is broken, and the rest follows.

What we are watching

• November 5, 2026: Q3 2026 earnings report. Verify that revenue landed inside the $255, $290 million guide, that 800G revenue grew approximately five times sequentially, and that initial 1.6T shipments are confirmed. Non GAAP gross margin versus the 29, 30.5% guide is the single most important line.
• October 2026: Any 8-K or 424B5 filing disclosing how much of the $600 million August ATM has been drawn and at what average price. This updates the dilution math in real time.
• September, October 2026: Watch for any customer qualification announcement or press release related to 1.6T transceivers beyond the existing anchor hyperscaler. A second named customer entering qualification would meaningfully de risk the 2027 revenue model.
• November 5, 2026: Management Q4 2026 guidance and any commentary on whether the $500 million plus Q4 revenue target is being maintained. This is the number that implies the $2 billion annualized run rate entering 2027.
• November, December 2026: Any update on CPO external light source qualification timelines, specifically whether the 400-milliwatt pump laser is in customer trials and when production gross margins are expected to reflect the 55, 65% range management has described.

The bottom line

The AI infrastructure wave is real, the 800G to 1.6T transceiver cycle is real, and $AAOI’s position inside that cycle is real. An 86% year over year revenue gain is not a rounding error. A $200 million single customer transceiver order is not a rounding error. The Amazon warrant anchoring the hyperscale relationship is not ambiguous. The demand story has already arrived.

What has not arrived is the evidence that the company can translate that demand into earnings per share that are worth something after accounting for the dilution required to build the factories. Three equity raises totaling more than a billion dollars in calendar 2026, with gross margins unchanged at 29.8%, is a warning that the business model has not yet proven it can self fund. The CPO margin story, if it arrives in 2027, changes everything. If it is delayed or misses expectations, shareholders who paid for the growth at current prices will have funded an impressive factory for someone else’s returns.

We would change our mind on the dilution concern if the Q3 print showed non GAAP gross margin moving above 31% and if management confirmed that the third $600 million ATM is being drawn conservatively, meaning well below its ceiling. A gross margin line that starts moving is the single observable that separates a compounding business from an expensive capital project.

Not financial advice.

References

1. Applied Optoelectronics Inc. Applied Optoelectronics Reports First Quarter 2026 Results. Applied Optoelectronics Investor Relations. Published May 7, 2026. Accessed September 9, 2026. https://investors.ao-inc.com/news-releases/news-release-details/applied-optoelectronics-reports-first-quarter-2026-results
2. Yahoo Finance / GuruFocus. Applied Optoelectronics Inc (AAOI) Q2 2026 Earnings Call Highlights. Yahoo Finance. Published August 7, 2026. Accessed September 9, 2026. https://finance.yahoo.com/markets/stocks/articles/applied-optoelectronics-inc-aaoi-q2-050943160.html
3. Zacks / Yahoo Finance. Applied Optoelectronics (AAOI) Surpasses Q2 Earnings and Revenue Estimates. Yahoo Finance. Published August 7, 2026. Accessed September 9, 2026. https://finance.yahoo.com/markets/stocks/articles/applied-optoelectronics-aaoi-surpasses-q2-023501524.html
4. Quartr. Applied Optoelectronics (AAOI) Investor Relations, Earnings Summary and Outlook. Quartr. Accessed September 9, 2026. https://quartr.com/companies/applied-optoelectronics-inc_5413
5. StockTitan. Applied Optoelectronics Q2 Earnings: $191.9M Revenue. StockTitan. Published August 6, 2026. Accessed September 9, 2026. https://www.stocktitan.net/news/AAOI/applied-optoelectronics-reports-second-quarter-2026-wjz4h91d9ngj.html
6. Finsee. Applied Optoelectronics Q2 2026 Earnings Review. Finsee. Published August 6, 2026. Accessed September 9, 2026. https://finsee.ai/earnings/aaoi/2026/q2/en/
7. Barchart. Applied Optoelectronics Returns With Guidance That Could Redefine Its AI Infrastructure Credibility. Barchart. Published August 5, 2026. Accessed September 9, 2026. https://www.barchart.com/story/news/3675338/applied-optoelectronics-returns-with-guidance-that-could-redefine-its-ai-infrastructure-credibility
8. Simply Wall St. Applied Optoelectronics Second Quarter Results, New Forecasts. Simply Wall St. Published August 9, 2026. Accessed September 9, 2026. https://simplywall.st/stocks/us/tech/nasdaq-aaoi/applied-optoelectronics/news/the-applied-optoelectronics-inc-nasdaqaaoi-second-quarter-re
9. ChartMill. AAOI Forecast, Price Target and Analyst Ratings. ChartMill. Accessed September 9, 2026. https://www.chartmill.com/stock/quote/AAOI/analyst-ratings
10. Investing.com. Applied Opt Stock Price Today, NASDAQ: AAOI. Investing.com. Accessed September 7, 2026. https://www.investing.com/equities/applied-opt
11. TradingView. NASDAQ:AAOI, Applied Optoelectronics Inc. TradingView. Accessed September 9, 2026. https://www.tradingview.com/symbols/NASDAQ-AAOI/
12. StockTitan / SEC 10-Q. Applied Optoelectronics Q2 2026 Quarterly Report (10-Q). StockTitan. Published August 6, 2026. Accessed September 9, 2026. https://www.stocktitan.net/sec-filings/AAOI/10-q-applied-optoelectronics-inc-quarterly-earnings-report-1b10770fa4b7.html
13. Applied Optoelectronics Inc. Form 8-K Q2 2026 Results (Exhibit 99.1). US Securities and Exchange Commission EDGAR. Published August 6, 2026. Accessed September 9, 2026. https://www.sec.gov/Archives/edgar/data/0001158114/000168316826006055/aaoi_ex9901.htm
14. Applied Optoelectronics Inc. Form 424B5, $600M ATM Program. US Securities and Exchange Commission EDGAR. Published May 14, 2026. Accessed September 9, 2026. https://www.sec.gov/Archives/edgar/data/0001158114/000110465926061146/tm2614157-1_424b5.htm
15. Applied Optoelectronics Inc. Form 424B5, $250M, $500M ATM Amendment. US Securities and Exchange Commission EDGAR. Published February, March 2026. Accessed September 9, 2026. https://www.sec.gov/Archives/edgar/data/1158114/000110465926020583/tm266210-1_424b5.htm
16. StockTitan. Applied Optoelectronics Expands ATM Program to $500M, Approximately 2.48M Shares Sold. StockTitan. Published March 12, 2026. Accessed September 9, 2026. https://www.stocktitan.net/sec-filings/AAOI/424b5-applied-optoelectronics-inc-prospectus-supplement-debt-securiti-0447d1e400a8.html
17. Globe and Mail / TipRanks. Applied Optoelectronics Establishes $600M ATM Program. Globe and Mail. Published May 15, 2026. Accessed September 9, 2026. https://www.theglobeandmail.com/investing/markets/markets-news/Tipranks/1966344/applied-optoelectronics-establishes-600-million-atm-program/
18. Yahoo Finance UK. Applied Optoelectronics (AAOI) Is Down 30.5% After Launching $600M ATM Equity Program. Yahoo Finance UK. Published August 2026. Accessed September 9, 2026. https://uk.finance.yahoo.com/news/applied-optoelectronics-aaoi-down-30-170816692.html
19. Simply Wall St. Why Applied Optoelectronics Is Down 6.9% After Launching $600M ATM. Simply Wall St. Published August 2026. Accessed September 9, 2026. https://simplywall.st/stocks/us/tech/nasdaq-aaoi/applied-optoelectronics/news/why-applied-optoelectronics-aaoi-is-down-69-after-launching
20. Malik Y. AAOI Stock Dips 10% Overnight After $600M Equity Offering Triggers Dilution Concerns. Stocktwits. Published August 23, 2026. Accessed September 9, 2026. https://stocktwits.com/news-articles/markets/equity/aaoi-stock-dips-10-overnight-after-600-m-equity-offering-triggers-dilution-concerns
21. TradingKey. AAOI Drops Over 12% Premarket as $600M ATM Sparks Dilution Concerns. TradingKey. Published August 24, 2026. Accessed September 9, 2026. https://www.tradingkey.com/analysis/stocks/us-stocks/262127665-aaoi-pre-market-drop-applied-optoelectronics-tradingkey
22. Tickeron. Applied Optoelectronics (AAOI) Shares Slide 21.7% in 30 Days on Dilution Worries. Tickeron. Published September 2, 2026. Accessed September 9, 2026. https://tickeron.com/blogs/applied-optoelectronics-aaoi-shares-slide-21-7-in-30-days-on-dilution-worries-16294/
23. Tickeron. Applied Optoelectronics (AAOI): Capacity Constraints Driving the AI Optics Opportunity. Tickeron. Published September 2026. Accessed September 9, 2026. https://tickeron.com/blogs/applied-optoelectronics-aaoi-capacity-constraints-driving-the-ai-optics-opportunity-16190/
24. Applied Optoelectronics Inc. AOI Introduces New 400mW Narrow Linewidth Pump Laser for Silicon Photonics and Co Packaged Optics. Applied Optoelectronics Investor Relations. Published December 18, 2025. Accessed September 9, 2026. https://investors.ao-inc.com/news-releases/news-release-details/aoi-introduces-new-ultra-high-power-semiconductor-laser-support
25. Fintel. AAOI Institutional Ownership. Fintel. Accessed September 9, 2026. https://fintel.io/so/us/aaoi
26. ChartMill. AAOI Stock Profile, Ownership. ChartMill. Accessed September 9, 2026. https://www.chartmill.com/stock/quote/AAOI/profile
27. MacroTrends. Applied Optoelectronics Shares Outstanding 2012, 2026. MacroTrends. Accessed September 9, 2026. https://www.macrotrends.net/stocks/charts/AAOI/applied-optoelectronics/shares-outstanding
28. FinanceCharts. Applied Optoelectronics Diluted Shares Outstanding. FinanceCharts. Published August 2026. Accessed September 9, 2026. https://www.financecharts.com/stocks/AAOI/income-statement/average-diluted-shares-outstanding
29. Simply Wall St. Applied Optoelectronics, Stock Analysis. Simply Wall St. Published September 6, 2026. Accessed September 9, 2026. https://simplywall.st/stocks/us/tech/nasdaq-aaoi/applied-optoelectronics
30. Simply Wall St. Applied Optoelectronics Reported Q2 Results, Is Valuation Too Rich? Simply Wall St. Published August 2026. Accessed September 9, 2026. https://simplywall.st/stocks/us/tech/nasdaq-aaoi/applied-optoelectronics/news/applied-optoelectronics-aaoi-reported-q2-results-and-new-gui
31. Applied Optoelectronics Inc. Applied Optoelectronics Closes Exchange of 2026 Notes and Concurrent Direct Offering. Applied Optoelectronics Investor Relations. Published December 23, 2024. Accessed September 9, 2026. https://investors.ao-inc.com/news-releases/news-release-details/applied-optoelectronics-closes-exchange-2026-notes-and
32. Quiver Quant. AAOI Insider Trading Data. Quiver Quant. Accessed September 9, 2026. https://www.quiverquant.com/news/Applied+Optoelectronics+Inc.+Unveils+New+400mW+Narrow-Linewidth+Pump+Laser+for+AI+Data+Centers+and+Co-Packaged+Optics
33. Yahoo Finance. Applied Optoelectronics Inc (AAOI), Stock Quote. Yahoo Finance. Accessed September 9, 2026. https://finance.yahoo.com/quote/AAOI/