MaxLinear
A California chip designer whose chips go into home internet boxes, Wi-Fi routers and, since 2025, the optical links between AI servers.
The scores
Growth3/4 ↑
Revenue up 29.7% in 2025, 55% in Q2 2026 and 72% in Q3, all organic, but the record is under two years old.
Business quality3/4 ↑
Design wins on power give real share gains, but the position is one generation old, unconfirmed from outside, and each new speed is competed again.
Financial position2/4 ↑
Cash of $81.1M is below Silicon Motion's $160M claim, $40M of the revolver is drawn, and the $125M term loan falls due in June 2028.
Execution record3/4 ↑
The optical forecast rose at every print, from about $100M to $250M-$270M, and Q3 landed in its guide, but Rushmore slipped into 2027.
Per share capture2/4 →
Stock pay is 15.6% of revenue and diluted shares rose from 73.1M in 2020 to 97.3M; one $20M buyback offsets little.
Potential2/4 →
A large, fast market, but the share rests on purchase orders, Rushmore winning 1.6T, and DSPs surviving linear-drive optics.
Risk profile2/4 →
An undecided arbitration claim larger than the cash resolves in one ruling; three-quarters of products ship to Hong Kong, China and Vietnam.
Valuation1/4 ↓
The price needs about $1.8B of revenue at a 30% GAAP margin: 1.6 times the 2022 record at nearly twice its margin.
Evidence quality2/4 →
The statements are filed, but optical revenue, the power edge, the three-player claim and Rushmore's timing are management's word, unconfirmed outside.
If it works 2.2× · if it fails you keep about 30%.
The Verdict
We rate MaxLinear unattractive. Its new optical chips are winning real orders at AI data centres, and the September quarter beat its forecast, but the other half of the company, broadband and Wi-Fi, is cyclical and fell by two-thirds two years ago, so the business is mixed. The price is priced for perfection: it needs sales about 1.6 times the company's best year, at nearly twice that year's profit margin. As a bet it is priced for success: about one and a half times the money if optical keeps growing with its market, under a quarter kept if it stalls. The next chip generation has already slipped once.
The long caseAI data centres keep adding faster optical links, and if MaxLinear's lower-power DSPs keep winning sockets at each new speed, its optical sales keep compounding.
The evidence under this - 7 claims
- Infrastructure also holds wireless backhaul and analog chips, so its 210% is not all optical; the filing gives no optical split (C01).
- The GAAP operating profit is about $30M on $218.1M of revenue; stock pay, performance equity and acquisition costs of about $34M make up the gap to the non-GAAP figure, and nine-month operating cash flow is negative.
- No figure for optical revenue is filed: the 10-Q reports only the Infrastructure category, which also holds wireless backhaul and analog chips.
- The shares fell 21.5% the day after the July raise, which suggests the market had already priced more.
- 2026-10-22editedFourth raise read on the 22 October call (S26): the range moved from $210M-$230M to $250M-$270M.
- 2026-10-22editedThird dating added from the 22 October call (S26): the first shipments are now expected earlier than July's date but later than January's.
- A 6% yield is a judgement; at 5% the profit needed falls to about $442M, or $1.5B of revenue at 30%.
- Measured on management's non-GAAP basis, which leaves out stock pay, a 35% margin would need about $1.5B of revenue.
- The filed rivals trade at a median of about 24 times sales, so the market prices the whole group this way (S16).
- 2026-10-22editedRe-struck on the C20 enterprise value of $6.98B: the revenue the price needs falls from about $2.4B to about $1.8B because the price fell 28%, not because the business earned more.
The company
MaxLinear designs chips and has them made by TSMC and UMC. Its fastest-growing chip, Keystone, sits inside the plug-in optical modules that link servers in AI data centres: it cleans up the fast electrical signal before a laser turns it into light.
Module makers and the cloud companies behind them buy it, management says, because it uses almost 40% less power than rival chips from Broadcom and Marvell. No outside source confirms that yet. The rest of the company sells chips for internet boxes, Wi-Fi routers and industrial equipment.
The evidence under this - 5 claims
- Credo, which MaxLinear's own 10-K names as a competitor, also sells optical DSPs to the same module makers (S21).
- LightCounting's public summaries and Cignal AI's vendor list do not name MaxLinear; no source outside the company states its share or its power advantage (C33).
- No figure for optical revenue is filed: the 10-Q reports only the Infrastructure category, which also holds wireless backhaul and analog chips.
- The shares fell 21.5% the day after the July raise, which suggests the market had already priced more.
- 2026-10-22editedFourth raise read on the 22 October call (S26): the range moved from $210M-$230M to $250M-$270M.
How it makes money
MaxLinear sells chips one purchase order at a time, mostly to equipment and module makers, with nearly half going through distributors. In Q2 2026 half of sales came from Infrastructure, the optical chips plus radio and analog parts, and half from broadband, Wi-Fi and industrial chips.
No customer signs a long contract. What keeps money coming is that a chip, once designed into a product, usually stays for that product's life, two to seven years.
The evidence under this - 4 claims
- 2026-10-22editedRe-sourced to the Q3 2026 10-Q (S24): concentration eased further; the Q2 geography split was dropped as no longer current.
The 10-K says customers buy on purchase orders, have cancelled or deferred them before, and that forecasts have been materially wrong (C29).
Where the money comes from, stream by stream
| Stream | Share of sales (Q2 2026) | How it earns it |
|---|---|---|
| Infrastructure | 50.4% | Keystone optical DSPs for data-centre modules, plus wireless backhaul and high-performance analog chips, sold to module makers and equipment makers. |
| Broadband | 26.6% | Chips for cable, fibre and DSL gateways, sold to box makers serving cable and telecom operators. |
| Connectivity | 14.2% | Wi-Fi and Ethernet chips for routers, gateways and switches. |
| Industrial and multi-market | 8.9% | Interface, power and storage-acceleration chips for industrial and other equipment, often through distributors. |
Who runs it
Kishore Seendripu co-founded MaxLinear in 2003 and has run it since. Steven Litchfield has been finance chief since 2018. In April the chief executive owned 6.0% and insiders 7.7% together. Long-term pay rides on sales and an adjusted profit measure, not on profit after stock pay or per share.
In August the chief executive sold about $50M of stock, about a ninth of his stake, outside any pre-arranged plan.
The evidence under this - 3 claims
He still held most of his stake, and a founder selling after a sixfold rise may be diversifying rather than signalling.
Why it wins, and who could beat it
Why customers come back
Once a module maker builds Keystone into a product and its cloud customer approves it, switching chips means redesigning and re-testing, so the chip tends to stay for years. That is MaxLinear's hold on its customers.
It is a hold on one generation. Each faster speed is a new design contest, so winning 800G links does not win 1.6T. And outside analysts do not name MaxLinear at all, so its share and its power advantage rest on management's word.
The evidence under this - 5 claims
- Management says Keystone's success is the foundation for multigenerational engagements, and a supplier already qualified at a cloud operator may be first in line for the next generation (S07).
- Keystone itself began ramping only in 2025 and won share against two incumbents, so the company has shown it can win a generation.
- Credo, which MaxLinear's own 10-K names as a competitor, also sells optical DSPs to the same module makers (S21).
- LightCounting's public summaries and Cignal AI's vendor list do not name MaxLinear; no source outside the company states its share or its power advantage (C33).
The threat
In optical chips MaxLinear fights Broadcom and Marvell, the incumbents, and Credo. The larger threat is a design that drops this chip altogether: NVIDIA already sells switches with the optics built in, and LightCounting, an industry research firm, expects such designs to slow sales of chips like Keystone from 2027. In broadband, Broadcom, Realtek and MediaTek sell the same kinds of chips.
The evidence under this - 6 claims
- Credo, which MaxLinear's own 10-K names as a competitor, also sells optical DSPs to the same module makers (S21).
- LightCounting's public summaries and Cignal AI's vendor list do not name MaxLinear; no source outside the company states its share or its power advantage (C33).
- Cignal says nearly all 1.6T modules still ship in configurations that use a DSP, and that 800G demand shows no sign of slowing (C26).
- LightCounting expects fully retimed transceivers, the kind that use a DSP, to contribute the largest dollar growth over its forecast (S17).
The rivals: AVGO, AVGO, MRVL, CRDO, NVDA, SMTC, MTSI, QCOM, SWKS, TXN, ADI, MCHP, and 1 more that is private, in-house or listed abroad.
Who else does this, rival by rival
| Ticker | Rival | Where it competes | How far along | Priced against it |
|---|---|---|---|---|
| AVGO | Broadcom PAM4 DSPsPriced against it: Shown, not counted | Infrastructure | On sale | Shown, not counted |
| Broadcom broadband and Wi-Fi chips | Broadband, Wi-Fi and industrial | On sale | ||
| MRVL | Marvell PAM4 DSPsPriced against it: Yes | Infrastructure | On sale | Yes |
| CRDO | Credo optical DSPsPriced against it: Yes | Infrastructure | On sale | Yes |
| NVDA | NVIDIA co-packaged optics and linear-drive modulesPriced against it: Small part of it | Infrastructure | On sale | Small part of it |
| Realtek and MediaTekPriced against it: Listed abroad | Broadband, Wi-Fi and industrial | On sale | Listed abroad | |
| SMTC | SemtechPriced against it: Yes | - | On sale | Yes |
| MTSI | MACOM Technology Solutions HoldingsPriced against it: Yes | - | On sale | Yes |
| QCOM | QUALCOMM IncorporatedPriced against it: One segment only | - | On sale | One segment only |
| SWKS | Skyworks SolutionsPriced against it: One segment only | - | On sale | One segment only |
| TXN | Texas Instruments IncorporatedPriced against it: One segment only | - | On sale | One segment only |
| ADI | Analog DevicesPriced against it: One segment only | - | On sale | One segment only |
| MCHP | Microchip TechnologyPriced against it: One segment only | - | On sale | One segment only |
How big it could get
How big this could get
Cignal AI, an industry research firm, counted almost 20 million fast optical modules shipped in Q1 2026 alone, and most need a chip like Keystone. LightCounting puts 2026 sales of chips for the newest 1.6T modules above $2B.
Against that, MaxLinear expects $210M-$230M of optical sales this year, at what management's own figures suggest is about $20-$25 a chip. The market is large; MaxLinear's slice of it is not public.
The evidence under this - 4 claims
- The chief executive rejected the suggestion that MaxLinear wins by pricing low, saying performance decides (S05).
- Units and revenue were given separately and may not describe the same mix of 400G and 800G parts.
- No figure for optical revenue is filed: the 10-Q reports only the Infrastructure category, which also holds wireless backhaul and analog chips.
- The shares fell 21.5% the day after the July raise, which suggests the market had already priced more.
- 2026-10-22editedFourth raise read on the 22 October call (S26): the range moved from $210M-$230M to $250M-$270M.
Who has signed
Nobody has signed a long contract. MaxLinear has design wins at cloud companies in the US and Asia, through module makers, and about 28 weeks of orders ahead. Orders can be cancelled, and management says a handful of customers will drive most of the volume.
The evidence under this - 4 claims
- Credo, which MaxLinear's own 10-K names as a competitor, also sells optical DSPs to the same module makers (S21).
- LightCounting's public summaries and Cignal AI's vendor list do not name MaxLinear; no source outside the company states its share or its power advantage (C33).
The 10-K says customers buy on purchase orders, have cancelled or deferred them before, and that forecasts have been materially wrong (C29).
- In January the same company expected Rushmore production revenue to start at the end of 2026 (C11).
- Nearly all 1.6T optics ship today in one configuration, and Broadcom and Marvell already supply the leading module makers (C26).
- 2026-10-22editedThe 22 October call (S26) puts the first Rushmore shipments in the first half of 2027, a quarter or two earlier than July's second-half ramp; a management claim, not yet filed.
Where the growth is coming from, business by business
Where the growth is coming from
| Business | Share of revenue | Growth: last quarter | Last half-year | Last year | Share in 3 years at its half-year rate |
|---|---|---|---|---|---|
| Infrastructure | 57% | +209.6% | +179.7% | +151.1% | 96% |
| Broadband, Wi-Fi and industrial | 43% | +8.2% | +10.5% | +17.3% | 4% |
| Whole company | 100% | +72.4% | +64.4% | +57.8% | 100% |
On the filed split through 2026-09-30, if each business keeps its last half-year's growth, the company grows about 107% next year and 162% in year 3, against 50.5% over the last twelve months. Infrastructure carries the case: it grows 180% against 10% for the rest, and would be 96% of revenue by then. A scenario, not a forecast: no business holds its rate.
How we read it: The score believes the business carrying the case: Infrastructure's 145% rests on Keystone ramps at a handful of customers, with Broadband down 6%. Under two years of record, so 3 either way.
The results
The numbers
Sales grew 56% over the twelve months to Sep 2026, faster than the 30% in the year to Dec 2025.
Operating margin is −0.9%, up from −5.5% in the year to Dec 2023.
Running costs take 59¢ of each sales dollar, up from 48¢ in the year to Dec 2021.
Shareholders' money lost 5.9% a year, down from 9.5% in the year to Dec 2021.
It used $20.6M of cash over the last twelve months, so the $81.1M it holds lasts about 3.9 years at that pace.
It owes $83.0M more than the cash it holds, and its operations bring in no cash to repay it.
The share count rose 23% in 4.8 years, and pay in shares costs 13% of sales.
All the numbers
Filed figures from the company's 10-K and 10-Q, read 2026-10-22. Fiscal years end in December.
| Dec 2021 | Dec 2022 | Dec 2023 | Dec 2024 | Dec 2025 | 12 mo to Sep 2026 | |
|---|---|---|---|---|---|---|
| Size and growth | ||||||
| Sales | $892M | $1.12B | $693M | $361M | $468M | $661M |
| Sales growth | +86% | +26% | −38% | −48% | +30% | +56% |
| Profit | ||||||
| Gross margin | 56% | 58% | 56% | 54% | 57% | 58% |
| Operating margin | 7.3% | 16% | −5.5% | −62% | −27% | −0.9% |
| Net margin | 4.7% | 11% | −11% | −68% | −29% | −4.4% |
| What growth costs, per sales dollar | ||||||
| Research and development | 31% | 26% | 39% | 62% | 45% | 34% |
| Selling, general and admin | 17% | 15% | 19% | 38% | 34% | 25% |
| Pay in shares | 6.7% | 7.3% | 8.0% | 18% | 16% | 13% |
| Capital spending | 4.4% | 3.7% | 1.9% | 4.9% | 2.7% | 2.0% |
| Cash | ||||||
| Cash from operations | $168M | $389M | $43.4M | −$45.3M | $19.6M | −$7.25M |
| Free cash flow | $129M | $347M | $29.9M | −$63.0M | $7.02M | −$20.6M |
| Balance sheet | ||||||
| Cash and short-term investments | $131M | $206M | $187M | $119M | $72.8M | $81.1M |
| Debt | $306M | $122M | $122M | $123M | $124M | $164M |
| Net cash | −$176M | $84.1M | $64.9M | −$4.42M | −$50.8M | −$83.0M |
| Returns and shares | ||||||
| Return on capital | 7.4% | 16% | - | - | - | - |
| Shares (diluted) | 79.7M | 80.9M | 80.7M | 83.6M | 86.6M | 97.7M |
What growth costs
Growth is making the company cheaper to run per sale: research fell from 63 cents of every sales dollar in 2024 to 38 cents over the last year. Stock paid to staff still costs about 16 cents per dollar, and factories cost little, because others make the chips.
The evidence under this - 2 claims
The last report (23 Jul)
Quarter ended June 2026, filed figures.
Sales for Q3 2026 were $218.1M, up 72% and inside the $210M-$220M guide, above its midpoint. Infrastructure, led by optical chips, rose 210% to $124.9M and is now 57% of sales, while broadband fell 8%. Counting the stock it pays staff, the company made a 13.9% operating profit; on its own adjusted measure the margin was 29.5%.
Management raised its 2026 optical forecast to $250M-$270M and guided Q4 sales to $240M-$260M. Cash from operations was negative $13.6M, because it paid chip makers ahead for 2027 orders.
What it did to the thesisIt met the test we set: sales above the guide's midpoint, the optical forecast raised and a profit after stock pay. It did not change what the price asks, which is sales far beyond the company's best year at a margin it has never earned.
The quarter in detail
Sales growth from the previous quarter, filed figures.
Earnings per share $0.35 against $0.33 expected; $0.02 as filed, the gap being what the company adds back.
The evidence under this - 4 claims
- Infrastructure also holds wireless backhaul and analog chips, so its 210% is not all optical; the filing gives no optical split (C01).
- The GAAP operating profit is about $30M on $218.1M of revenue; stock pay, performance equity and acquisition costs of about $34M make up the gap to the non-GAAP figure, and nine-month operating cash flow is negative.
The guide was 27% above Q2 revenue and the quarter landed 29% above it, at the top half of the range; the beat over the vendor's estimate was 1.4% (C22).
- 2026-10-22editedThe guided quarter was reported: management's forecast tested against the filed Q3 figures (S24) and becomes a fact.
The guide is 15% above Q3 revenue at the midpoint; Q3's revenue beat the vendor estimate by 1.4%, a smaller beat than the 2.6% of Q2 that preceded a 21.5% fall (C22).
What management did with the money
In 2023 MaxLinear agreed to buy Silicon Motion, then walked away. Silicon Motion is still pursuing it in arbitration for a $160M fee plus damages, more than the company's cash. Since then it has paid down debt to $125M and bought back $20M of stock, but drew $40M on its credit line this year to pay for chips ahead.
Insiders in 2026Q3: 1 purchases and 18 sales, from their Form 4 filings.
The evidence under this - 4 claims
- 2026-10-22editedRead again in the Q3 10-Q (S24): still confidential and undecided, with no date; no change to the claim.
- 2026-10-22editedRe-sourced to the Q3 10-Q (S24) and the Q4 guide (S26): shares up from 90.7M to 91.0M.
- 2026-10-22editedRe-struck to the Q3 2026 10-Q (S24) and the 16 October amendment (S25): cash up $16.3M, the revolver drawn $40.0M and enlarged, prepayments up.
The price, and where we disagree
In short. Priced for perfection. The price assumes MaxLinear grows far past its best-ever year and earns a profit margin it has never earned.
The shares fell 28% in the two weeks to 21 October, from $105.49 to $75.79, before the results. They had risen from $16 to over $100 in a year, so the price swings hard on mood.
The 22 October results met the test we set. The next tests are the Q4 results in February and whether the first 1.6T chip sales arrive in the first half of 2027, as management now says. LightCounting wrote in September that optical market growth must be at its peak.
The evidence under this - 4 claims
- Infrastructure also holds wireless backhaul and analog chips, so its 210% is not all optical; the filing gives no optical split (C01).
- The GAAP operating profit is about $30M on $218.1M of revenue; stock pay, performance equity and acquisition costs of about $34M make up the gap to the non-GAAP figure, and nine-month operating cash flow is negative.
- In January the same company expected Rushmore production revenue to start at the end of 2026 (C11).
- Nearly all 1.6T optics ship today in one configuration, and Broadcom and Marvell already supply the leading module makers (C26).
- 2026-10-22editedThe 22 October call (S26) puts the first Rushmore shipments in the first half of 2027, a quarter or two earlier than July's second-half ramp; a management claim, not yet filed.
How we read the multiple
Buyers pay about 10.6 times the last year's sales, down from 16.9 on 5 October. The best year MaxLinear ever had was 2022: $1.12B of sales and 16 cents of operating profit on each dollar.
To earn 6% a year after tax on today's price, a common rule of thumb, MaxLinear must:
- Make about $530M a year of operating profit, after stock pay.
- Sell about $1.8B a year, keeping 30 cents of profit on each dollar.
- That is about 1.6 times its best year, at nearly twice its best margin.
Against its own history. Far dearer than its own past, when it was a broadband chip company; the higher price is a bet on the optical business, not something it has earned.
The evidence under this - 4 claims
- On the vendor's share count the multiple would be a few tenths higher.
- Counting the $28.8M of restricted cash would lower the enterprise value by under 1%.
- 2026-10-22editedRe-struck at the 21 October close ($75.79, from $105.49 on 5 October) with the Q3 filing's share count, debt, cash and trailing revenue (S24): 10.6 times from 16.9.
- A 6% yield is a judgement; at 5% the profit needed falls to about $442M, or $1.5B of revenue at 30%.
- Measured on management's non-GAAP basis, which leaves out stock pay, a 35% margin would need about $1.5B of revenue.
- The filed rivals trade at a median of about 24 times sales, so the market prices the whole group this way (S16).
- 2026-10-22editedRe-struck on the C20 enterprise value of $6.98B: the revenue the price needs falls from about $2.4B to about $1.8B because the price fell 28%, not because the business earned more.
The GAAP record is a 16.1% operating margin, in 2022 (C04), and stock pay runs at 15.6% of revenue, which the non-GAAP basis leaves out (C14).
- 2026-10-22editedQ3 actuals set beside the targets (S26): non-GAAP gross margin 60.0% against 65%, non-GAAP operating margin 29.5% against 30%-35%; the targets were not restated or withdrawn on the call.
Priced against its rivals
The competitors its own annual report names, each confirmed against the rival's filing. Measured on company value for each $1 of sales, cheapest first:
| Ticker | Company | company value for each $1 of sales |
|---|---|---|
| SMTC | Semtech | 15.8× |
| MXL | MaxLinear | 17.3× |
| MTSI | MACOM Technology Solutions Holdings | 22.9× |
| CRDO | Credo Technology Group Holding | 25.9× |
| MRVL | Marvell Technology | 27.2× |
| AVGO | Broadcom not in the median | 20.5× |
The middle of the other 4 is 24.4. Broadcom Inc. competes with it in one product line inside a much bigger company, so its price says little about this business and it is not counted in the middle. MaxLinear, Inc. is 2nd cheapest of 5.
Rivals' figures come from a data provider, so read the spread of the group rather than any single pair.
Cheaper than its optical rivals, but the whole group has been re-priced for AI, and half of MaxLinear is slower broadband.
Private or foreign rivals that cannot be priced: Realtek Semiconductor Corp., MediaTek, Inc., Renesas Electronics Corporation, vertically integrated module makers.
The bet: what it pays if it works
This bet is on one business: Infrastructure, mostly the optical chips. Broadband, Wi-Fi and industrial are counted separately, at the modest price listed chip makers in those markets fetch for each dollar of sales.
If Infrastructure grows with its market from today's pace, valued like its optical rivals, the stock is worth only modestly more than today: an expensive bet. If optical stalls, shareholders keep mostly the slower half, a small part of today's price.
| Infrastructure revenue in two to three years | PAM4 DSP chips needed | Valued like its peers | Like the cheapest |
|---|---|---|---|
| $234.72M | about 10.4M at 100% kept | 1.0× | 0.7× |
| $598.50M (the big win) | about 26.6M at 100% kept | 2.2× | 1.5× |
| $630.00M | about 28.0M at 100% kept | 2.3× | 1.5× |
To pay 2× (a Fair bet) it would need about $549.18M of Infrastructure revenue (about 24.4M PAM4 DSP chips); to pay 5× (a Cheap bet), about $1.46B of Infrastructure revenue (about 64.7M PAM4 DSP chips), at the same multiple.
Each row is Infrastructure alone, plus the rest of the company (Broadband, Wi-Fi and industrial) counted at about $1.34B.
The steps it has to take
- Done 2026 optical forecast raised from about $100M to $210M-$230M
- Done Q2 2026 revenue of $168.8M, up 55%, with Infrastructure half of it
- Done Q3 2026 revenue of $218.1M within the $210M-$220M guide
- In progress 2026 optical revenue of $250M-$270M delivered (raised in October from $210M-$230M) due 2027-01-31
- Slipped First Rushmore 1.6T production revenue due 2027-12
- Not started Washington and Annapurna volume ramps due 2028-12
- Not started A quarter with a GAAP operating margin of 25% or more due 2027-12
The detail behind the label
The evidence under this - 8 claims
None is a like-for-like broadband chip seller: Qualcomm is mostly handsets, Skyworks mostly radio parts, and Realtek and MediaTek, the closest rivals, list in Taiwan.
- 2026-10-22editedThird dating added from the 22 October call (S26): the first shipments are now expected earlier than July's date but later than January's.
- In January the same company expected Rushmore production revenue to start at the end of 2026 (C11).
- Nearly all 1.6T optics ship today in one configuration, and Broadcom and Marvell already supply the leading module makers (C26).
- 2026-10-22editedThe 22 October call (S26) puts the first Rushmore shipments in the first half of 2027, a quarter or two earlier than July's second-half ramp; a management claim, not yet filed.
- 2026-10-22editedRe-struck to the Q3 2026 10-Q (S24) and the 16 October amendment (S25): cash up $16.3M, the revolver drawn $40.0M and enlarged, prepayments up.
- 2026-10-22editedRead again in the Q3 10-Q (S24): still confidential and undecided, with no date; no change to the claim.
Where we disagree with the market
An AI optics winner whose optical sales triple this year and keep compounding through 2028.
The optical wins are real but cover one chip generation, in a market near its fastest growth, beside a cyclical broadband half.
The market is paying for several more years of fast growth and a much fatter profit margin. The evidence covers one generation: the next chip has slipped, and analysts expect the market's growth to slow from 2027.
We are wrong if quarterly sales reach $300M with a 25% operating margin after stock pay, and the 1.6T chip starts selling on time.
What analysts say, and the evidence
The evidence under this - 6 claims
- 2026-10-22editedRefresh date rolled: the panel's estimates and ratings predate the 22 October print and have not yet been updated by the vendor; figures unchanged.
- No figure for optical revenue is filed: the 10-Q reports only the Infrastructure category, which also holds wireless backhaul and analog chips.
- The shares fell 21.5% the day after the July raise, which suggests the market had already priced more.
- 2026-10-22editedFourth raise read on the 22 October call (S26): the range moved from $210M-$230M to $250M-$270M.
- A 6% yield is a judgement; at 5% the profit needed falls to about $442M, or $1.5B of revenue at 30%.
- Measured on management's non-GAAP basis, which leaves out stock pay, a 35% margin would need about $1.5B of revenue.
- The filed rivals trade at a median of about 24 times sales, so the market prices the whole group this way (S16).
- 2026-10-22editedRe-struck on the C20 enterprise value of $6.98B: the revenue the price needs falls from about $2.4B to about $1.8B because the price fell 28%, not because the business earned more.
- 2026-10-22editedThird dating added from the 22 October call (S26): the first shipments are now expected earlier than July's date but later than January's.
What could go wrong, and how we would know
The risks
- An arbitration bill. Silicon Motion seeks a $160M fee plus damages, more than MaxLinear's cash, in one undecided ruling.
- Optics without the chip. Designs that drop the DSP could slow sales of Keystone-type chips from 2027.
- The next generation slips. The 1.6T chip moved from late 2026 into 2027, and each speed is a new contest.
- Broadband falls again. This half of the company fell by two-thirds from 2022 to 2024.
- Shipped through Asia. Three-quarters of products ship to Hong Kong, China and Vietnam, exposed to trade rules.
- The founder is selling. The chief executive sold about $50M of stock in August, outside a pre-arranged plan.
The evidence under this - 6 claims
- 2026-10-22editedRead again in the Q3 10-Q (S24): still confidential and undecided, with no date; no change to the claim.
- 2026-10-22editedThird dating added from the 22 October call (S26): the first shipments are now expected earlier than July's date but later than January's.
- 2026-10-22editedRe-sourced to the Q3 2026 10-Q (S24): concentration eased further; the Q2 geography split was dropped as no longer current.
He still held most of his stake, and a founder selling after a sixfold rise may be diversifying rather than signalling.
What we still do not know
- MaxLinear's share of optical DSP chips, and what each sells for. Only a paid industry report says.
- When the arbitration will be decided, and for how much. The proceedings are confidential.
- How much of Infrastructure is optical. The company reports only the total.
The evidence under this - 3 claims
- 2026-10-22editedRead again in the Q3 10-Q (S24): still confidential and undecided, with no date; no change to the claim.
- No figure for optical revenue is filed: the 10-Q reports only the Infrastructure category, which also holds wireless backhaul and analog chips.
- The shares fell 21.5% the day after the July raise, which suggests the market had already priced more.
- 2026-10-22editedFourth raise read on the 22 October call (S26): the range moved from $210M-$230M to $250M-$270M.
What would change our view
- Unattractive→NeutralA cheaper price on an unchanged business
It needs:
- ○The company's total value falls to 8 times its last year of sales, with the business unchanged.about ≈$57.20 today, 27% below the last close
- Unattractive→NeutralThe next generation wins and the margin arrives
It needs all of:
- ○Quarterly sales of $300M or more.checked at the next results, due by 4 Feb
- ○A quarter with an operating profit of 25 cents per sales dollar, after stock pay.checked at the next results, due by 4 Feb
- ○The 1.6T chip selling in volume at two or more customers.watched as it happens
We would also be wrong if
- Optical sales for 2026 fall below $250M, the bottom of management's October range.
- LightCounting cuts its 2027 growth forecast for optical modules below 30%.
- The arbitration awards Silicon Motion more than MaxLinear's cash.
What is coming
- 4 February 2027 Q4 2026 results and the 2026 10-K: Revenue against the $240M-$260M guide, whether 2026 optical revenue landed in the raised $250M-$270M, the year's GAAP operating result and any update on the arbitration.
Every dated event ahead
- 4 February 2027 Q4 2026 results and the 2026 10-K: Revenue against the $240M-$260M guide, whether 2026 optical revenue landed in the raised $250M-$270M, the year's GAAP operating result and any update on the arbitration.
- June 2027 Rushmore 1.6T first production revenue: Whether the initial design win management says will ship in the first half of 2027 does, and whether MaxLinear wins the next speed's sockets or Keystone's position ends with 800G.
- 23 June 2028 Term loan maturity: Whether $125M is repaid from cash flow or refinanced.
Since we wrote this
- 22 OctEarningsQ3 beat and a fourth optical raise; Unattractive holdsSales of $218.1M beat the guide's midpoint and optical guidance rose to $250M-$270M, but the price still needs about $1.8B of revenue at a 30% margin.
- 15 OctBet moveBet re-struck: Priced for success to SpeculationRe-priced at the 14 October close; the rating moved from Expensive bet to Fair bet.
- 6 OctThesis writtenUnattractive, medium convictionReal optical chip wins, but the price needs twice the company's best year at a margin it has never earned.
Also checked 2 times with nothing new to act on.
Sources 24
Sources - 24
Every claim above cites one of these, or states the reasoning that derived it. The quote is frozen at the date it was read: a link can rot, and an agent fetch returns a reading of a page rather than the page. Backs counts the whole claim ledger, so it can name a claim this page does not surface.
- S01primarycompany filing2026-01-29business, revenue by category, competitors, customer concentration, design-in life cycles, purchase-order terms, headcount, risk factorsbacksC01C02C04C05C08C28C29C36
- S02primarycompany filing2026-01-29full-year 2025 figures and the Q1 2026 guidebacksC02
- S03primarycompany filing2026-07-23revenue by category, concentration, shipment geography, balance sheet, debt, share count, the Silicon Motion disputebacksC01C03C08C15C16C17C20C22C23
- S04primarycompany filing2026-07-23Q3 guidance and the non-GAAP reconciliationbacksC03C12C14
- S05primarytranscript2026-01-29Fourth quarter 2025 earnings call transcriptthe first 2026 optical forecast, the 4M to 6M unit figure and the pricing exchange, Rushmore's first date, lead times, the buyback
we expect Keystone to generate about $100 million to $130 million in revenue in '26 with potential upside along with a further step function increase in run rate as we move into 2027. ... We expect Rushmore production revenue ramp starting at the end of 2026. ... this year, we could do anywhere between 4 million to 6 million units of PAM4 transceivers ... As you know, we've got 28-week lead times.
backsC06C07C09C11C16C32 - S06primarytranscript2026-04-23First quarter 2026 earnings call transcriptthe second optical forecast and the first fibre design win at a cloud operator
we are increasing our expectations for 2026 optical data center revenue to $150 million to $170 million range.
backsC06 - S07primarytranscript2026-07-23Second quarter 2026 earnings call transcriptthe third optical forecast, Keystone's power claim, the three-player claim, Rushmore's new date, the handful of customers, the long-term margin targets; analyst questions from Stifel, Susquehanna, Northland and Benchmark
we are once again raising our expectations for 2026 optical data center revenue to be between $210 million to $230 million with continued growth as run rates expand into 2027. Keystone, our 100-gigabit per lane 5-nanometer CMOS PAM4 DSP and SerDes technology continues to ramp into high-volume production at major hyperscale customers across U.S. and Asia for 400-gig and 800-gig deployments, delivering almost 40% lower consumption in power than competition ... We expect Rushmore, our 1.6 terabit optical PAM4 DSP at 200 gigabit per lane speeds to become an important optical connectivity growth driver beginning in 2027 ... For both Annapurna and Washington, we expect initial revenue in 2027, followed by a more meaningful volume ramp in 2028. ... there are only 3 players right now, and we are one of them.
backsC06C07C09C10C11C12C13C15 - S10primarycompany filing2026-04-02biographies, beneficial ownership, pay design and 2025 paybacksC18C19C35
- S12primarycompany filing2026-10-06revenue, margins, cost lines, stock pay, cash flow and diluted shares by year and for the trailing four quartersbacksC04C14C15C16C20C22C34
- S13secondaryvendor data2026-10-06Filed revenue split re-keyed by FMP, frozen at content/MXL/segment-history.jsonrevenue by category for 2020 to 2025 and each quarter, checked against filed revenuebacksC05
- S14primarycompany filing2026-08-10Silicon Motion Technology Form 6-K, interim report (Exhibit 99.2) Silicon Motion Technologythe claimant's own account of the arbitration and what it seeksbacksC17
- S15primarycompany filing2026-08-18each sale's size, price and Rule 10b5-1 flag, and a year of insider transactionsbacksC18
- S16secondaryvendor data2026-10-06Research pack, frozen at content/MXL/research-pack.jsonprice, returns, decomposition, analyst panel and estimates, earnings history, FINRA short interest, the competitors:filed cohortbacksC20C22C23
- S17primaryindustry data2026-02-26the analyst's view of the DSP market, its 1.6T size and the threat from linear-drive optics; read through ir_fetch
In 2025, hyperscalers' massive investments in AI infrastructure caused 800G PAM4 chipset shipments to nearly triple, with sales more than doubling year over year. ... Sales of 1.6T chipsets will exceed $2 billion in 2026 and grow rapidly through 2029. ... fully retimed Ethernet transceivers will still contribute the largest dollar growth, followed by active cables and linear optics (LPO/LRO/CPO). ... We expect that growth in sales of PAM4 chipsets will moderate in 2027-2031, as volume deployments of linear-drive solutions (LPO and CPO) will negatively impact DSP sales.
backsC24C33 - S18primaryindustry data2026-09-30LightCounting newsletter - We Must Be at the Peak of This Super-cycle (Ethernet Optics report, September 2026) LightCountingthe market's growth path and the warning that a flat year follows each growth period; read through ir_fetch
We now expect 120% growth in the Ethernet optical transceiver market for 2026 ... Even in this scenario, we assume that the market growth rate will decline from 120% in 2026 to 76% in 2027 and 20% by 2031. We must be at the peak of this super-cycle in terms of growth rate. ... growth periods of 2-3 years are interrupted by a flat or negative year.
backsC25 - S19primaryindustry data2026-06-18Cignal AI press release - Datacom Optical Component Revenue Doubles to $7.7 Billion in 1Q26 Cignal AIan independent unit count for the modules the DSPs go into, and the CPO timetable; read through ir_fetch
High-speed (400G+) datacom module shipments set a record; up 145% year-over-year. ... Almost 20 million modules shipped in the quarter as AI build-outs accelerated. 800GbE now accounts for the majority of high-speed datacom module shipments, and unit shipments will more than double in 2026. ... 1.6TbE module shipments grew 50% sequentially and will exceed 10 million units in 2026. Nearly all 1.6TbE optics currently ship in DR (silicon photonics) configurations. ... CPO deployments are imminent with Nvidia, Broadcom, and TSMC advancing 200G/lane platforms.
backsC26C33 - S20primaryother2026-10-01the attacker's own account of what it replaces and who has adopted it; the quote frozen by the AAOI thesis
Replacing pluggable transceivers with silicon photonics on the same package as the ASIC, NVIDIA CPO innovations provide 5x better power efficiency ... Design simplicity and power efficiency drives CoreWeave, Lambda, Meta, Microsoft, and Oracle Cloud Infrastructure as first adopters. ... Spectrum-X Ethernet Photonics Ramps to Full Production
backsC27 - S21primarycompany filing2026-06-23Credo Technology Form 10-K for fiscal 2026 Credo Technology Groupa rival selling optical DSPs to the same buyers; its competition section names Broadcom, Marvell and Astera Labs and not MaxLinear
We sell our products to hyperscalers, Neoclouds, OEMs, ODMs, CMs and optical module manufacturers, as well as into the enterprise and HPC markets.
backsC33 - S22primarycompany filing2026-03-11Marvell Technology Form 10-K for fiscal 2026 Marvell Technologythe incumbent DSP seller's own filing; its 24-name competitor list does not include MaxLinear
We are a fabless supplier of high-performance semiconductor products with core strengths in developing and scaling complex System-on-a-Chip architectures
backsC33 - S23secondaryvendor data2026-10-06FMP key-metrics-ttm for SYNA, QCOM and SWKS, fetched 2026-10-06 Financial Modeling Preptrailing company-value-to-sales for listed sellers of Wi-Fi, broadband and radio chips, through scripts/api/fmp.py
SYNA evToSalesTTM 4.24 enterpriseValueTTM 5.08B; QCOM evToSalesTTM 4.54 enterpriseValueTTM 199.97B; SWKS evToSalesTTM 3.02 enterpriseValueTTM 12.13B
backsC37 - S24primarycompany filing2026-10-22Q3 and nine-month revenue by category, customer concentration, balance sheet, debt and the October credit amendment, share count, the Silicon Motion notebacksC08C12C15C16C17C20C38C39
- S25primarycompany filing2026-10-22the Item 1.01 amendment raising the revolver to $200M and the Item 2.02 resultsbacksC15C38
- S26primarytranscript2026-10-22Third quarter 2026 earnings call transcriptQ3 revenue and margins, the fourth optical raise, the Q4 guide, Rushmore's first-half 2027 design win, prepayments for 2027 wafers; analyst questions from Stifel, Loop, Stonex, Susquehanna, BNP Paribas, Needham, Northland and Wells Fargo
Reflecting this momentum, we are raising our 26 optical data server revenue outlook to $250 million to $270 million. ... We currently expect revenue in the fourth quarter of 26 to be between $240 million and $260 million ... there is an initial design win that will start shipping in the first half.
backsC06C10C11C12C13C15C16C39C40 - S27primarycompany filing2026-10-22Q3 revenue, GAAP and non-GAAP margins and earnings per share, the Q4 guidebacksC12C39C40