An analysis of AMD’s Q3 2026 results and Qualcomm’s fiscal Q4 2026 results (November 2026)
Introduction: two key units in the AI advance
In the army of Alexander the Great, the advance did not rely on numbers alone but on the combination of the heavy cavalry of the Companions (hetairoi) and the elite infantry of the hypaspists. The Companions delivered the decisive blow at the decisive moment, while the hypaspists provided flexibility and held the flanks, without which the main phalanx could not move forward.
In a similar way, today’s advance of artificial intelligence depends not only on aggregate capital spending but also on specific companies playing different roles. AMD acts like the Companions - its concentration in Data Center and the ramp of Instinct and EPYC make it a central player in the current AI infrastructure cycle. Qualcomm is closer to the hypaspists - it is expanding into automotive and edge markets and taking its first steps in data centers. The analogy has limits: Qualcomm’s Data Center business is still a target, not a force on AMD’s scale, and its core business - handsets - is under pressure. Qualcomm is therefore better seen as a diversification option on the cycle than as a proven second strike force.
AMD: Data Center revenue nearly doubles
AMD reported revenue of $13.377 billion for Q3 2026 (quarter ended September 26, 2026), up 45% year over year. Growth came almost entirely from Data Center, which reached a record $8.546 billion (+97% YoY) and now accounts for 63.9% of total revenue. Client and Gaming revenue fell to $3.722 billion (-8%): Client rose to $2.949 billion (+7%), while Gaming dropped to $773 million (-40%). Embedded grew to $1.109 billion (+29%).
For context: NVIDIA remains the dominant supplier of AI accelerators, and AMD is its main challenger. AMD’s strike at the center is strong in growth rate, but it is delivered from the position of the number two player, not the market leader.
AMD Q3 2026: revenue by segment
| Segment | Revenue Q3 2026 | YoY change | Share |
|---|---|---|---|
| Data Center | $8.546B | +97% | 63.9% |
| Client and Gaming | $3.722B | -8% | 27.8% |
| of which Client | $2.949B | +7% | - |
| of which Gaming | $0.773B | -40% | - |
| Embedded | $1.109B | +29% | 8.3% |
| Total | $13.377B | +45% | 100% |
Operating income and quality of earnings
Segment operating income shows the same concentration. Data Center generated $2.882 billion of operating income at a 33.7% operating margin (calculated from reported absolute figures), Client and Gaming $573 million (15.4%), and Embedded $453 million (40.8%). Total GAAP operating income was $2.832 billion, and non-GAAP operating income about $3.9 billion.
The gap is largely due to the non-cash change in the fair value of customer warrants, reflected in the “Other” line (-$1.076 billion together with corporate costs) and excluded from non-GAAP. The warrants are tied to strategic agreements with large AI customers. They improve visibility into future revenue, but they come at a price: potential shareholder dilution and higher customer concentration. Dilution on full exercise is real, even if spread over time.
AMD Q3 2026: operating income by segment
| Segment | Operating income | Operating margin |
|---|---|---|
| Data Center | $2.882B | 33.7% |
| Client and Gaming | $0.573B | 15.4% |
| Embedded | $0.453B | 40.8% |
| Other (warrants and corporate) | -$1.076B | - |
| Total (GAAP) | $2.832B | 21.2% |
| Total (non-GAAP) | ~$3.9B | ~29% |
Note: the “Other” line includes the fair value change of customer warrants and unallocated corporate costs.
Free cash flow and capital expenditure
Capital expenditure for the quarter was $648 million. Free cash flow, however, shrank to $554 million, mainly due to inventory build and investment in supply chain capacity (wafers, substrates, advanced packaging and memory). The company ended the quarter with $15.8 billion in cash and short-term investments.
Conversion in this article means the share of non-GAAP operating income that turns into free cash flow in the period (free cash flow / non-GAAP operating income). For Q3 2026 it is about 14% ($554 million / about $3,900 million).
AMD Q3 2026: cash flow and liquidity
| Metric | Q3 2026 | Note |
|---|---|---|
| Capital expenditure | $0.648B | Funded from operating cash flow and cash |
| Free cash flow | $0.554B | Conversion ≈ 14% of non-GAAP op. income |
| Cash and short-term investments | $15.8B | Strong liquidity |
The following scenarios are the author’s assumptions, not company guidance. Free cash flow could start converging toward non-GAAP operating income in the second half of 2027 if, at the same time: inventories stabilize or decline after the capacity build peak; Data Center revenue keeps growing above 50% YoY; and gross margin stays in the 54-56% range (Q3 2026: 54% GAAP, 56% non-GAAP). Under these conditions conversion could return to roughly 40-60%, assuming the net change in working capital normalizes to about 5-8% of revenue and capex runs at about 4-6% of revenue.
Sensitivity is significant. If working capital stays around 10% of revenue, conversion falls below 40% even with stable inventories. If memory and packaging make up about 40% of Instinct cost of goods instead of the assumed 25-35%, Data Center gross margin compression could exceed 5 percentage points.
In a middle scenario - inventories stabilize but gross margin contracts by about 3 percentage points - conversion likely stays in the 25-40% range. That pushes convergence out to 2028. If inventories keep rising while margins compress, the risk-reward deteriorates further.
AMD: 2027 conversion scenarios (author’s assumptions)
| Scenario | Conditions | Conversion (FCF / non-GAAP op. income) | Valuation implication |
|---|---|---|---|
| Full | Stable inventories, DC above 50%, gross margin 54-56% | 40-60% | Valuation easier to justify |
| Middle | Stable inventories, gross margin ~-3 pp | 25-40% | Valuation harder to justify |
| Weak | Rising inventories, margin compression | below 25% | Worse risk-reward |
Qualcomm: automotive nearly doubles
Qualcomm reported non-GAAP revenue of about $10.44 billion for fiscal Q4 (quarter ended September 27, 2026), down about 7% YoY. QCT (chips) contributed about $9.0 billion and QTL (licensing) about $1.3 billion. Within QCT, Handsets fell to $5.14 billion (-26%), Automotive reached a record $1.99 billion (+89%), and IoT rose to $1.89 billion (+4%).
Two Automotive figures must be kept clearly apart:
- Full fiscal 2026: about $6.0 billion (+52%) - reported annual revenue;
- Q4 run-rate: about $8 billion (1.99 x 4) - the quarter annualized, not a reported annual result.
The increase in Automotive and IoT offset about 56% of the dollar decline in Handsets for the quarter. Diversification is growing, but it does not yet stabilize the company.
Qualcomm fiscal Q4 2026: revenue by segment
| Segment / subsegment | Revenue FQ4 2026 | YoY change | Note |
|---|---|---|---|
| QCT (total) | ~$9.0B | ~-8% | Chips |
| Handsets | $5.14B | -26% | Main decline |
| Automotive | $1.99B | +89% | FY2026: $6.0B (+52%); Q4 run-rate ~$8B |
| IoT | $1.89B | +4% | |
| QTL | ~$1.3B | ~-7% | Licensing |
| Total (non-GAAP) | ~$10.44B | ~-7% |
What growth Qualcomm needs to stabilize
Illustrative math: if Handsets keep declining 15-20% a year, on an annual run-rate of about $20.5 billion (5.14 x 4) that is a loss of roughly $3-4 billion of revenue per year. To stabilize total revenue, Automotive and Data Center need to add about the same amount.
- Against a run-rate base of about $8 billion in Automotive, that means combined growth of about 40-50%.
- Against reported FY2026 Automotive of about $6.0 billion, the required growth is about 50-70% - a steeper path.
The 40-50% range is an illustration by the author against the run-rate base, not a forecast. Qualcomm’s target of $5 billion in Data Center revenue for fiscal 2027, supported by its AI data center collaboration with Amazon, would cover much of the required increase if delivered. A significant part of automotive growth is backed by design wins already secured; the rest requires new demand.
Risk-reward comparison
AMD is the preferred position if free cash flow converges toward operating income in the second half of 2027 (conversion of 40% or more) and margin compression stays within 2-3 percentage points. At the regular-session close on November 5, 2026, AMD was $404.40 (-17.3% on the day) and Qualcomm $156.75 (-9.5%). AMD’s sharper repricing improves the relative entry versus Qualcomm, but it does not remove the conversion condition and it is not an automatic buy. Qualcomm becomes preferred if AMD lands in the middle scenario (conversion of 25-40%) while Qualcomm adds about $3-4 billion from Automotive and Data Center. With partial execution at both, Qualcomm offers better protection thanks to its steadier current cash flow and more moderate valuation (trailing P/E about 12.3x at the same close).
Prices at the November 5, 2026 close and an author framework for multiples. AMD $404.40 (-17.3% versus the prior day), Qualcomm $156.75 (-9.5%; trailing P/E about 12.3x). Source: Yahoo Finance, regular session. AMD’s session decline already exceeds the earlier illustration of an entry about 10-15% lower, so that recommendation is dropped. Relative repricing favors AMD versus Qualcomm, but free-cash-flow conversion remains the decisive filter. With full execution, a P/E premium for AMD of about 6-10 points over Qualcomm can still be considered justified. In the middle scenario, further AMD multiple compression of about 8-12 points versus levels before November 5, 2026 would equalize expected returns with Qualcomm. A staged AMD entry fits only investors who accept execution and dilution risk, with adds contingent on inventory stabilization and progress toward conversion of 40% or more.
Which position is preferred (author’s framework)
| Scenario | AMD | Qualcomm | Preferred |
|---|---|---|---|
| Full execution | Conversion 40%+, compression up to 2-3 pp | +$3-4B from Auto and DC | AMD |
| AMD middle + Qualcomm success | Conversion 25-40% | +$3-4B from Auto and DC | Qualcomm |
| Partial at both | Conversion below 25%, compression | Increase below $3B | Qualcomm |
Conclusion: two different roles in the AI advance
AMD currently delivers the main blow through its Data Center concentration and its ability to turn demand for accelerators and server CPUs into fast revenue growth. Qualcomm widens the front through an automotive business that is already real and through the start of a data center ramp that must offset weakness in handsets.
The two companies are not interchangeable. If AMD fails to turn growth into sustainable free cash flow in the second half of 2027, the strike at the center weakens. If Qualcomm fails to add the required $3-4 billion from Automotive and Data Center, the widening of the front slows. The AI advance can continue without them, but not with the same speed, reach or resilience.
Frequently asked questions
What was the main driver of AMD’s growth in Q3?
Data Center: $8.546 billion (+97%), 63.9% of total revenue.
Why is AMD’s free cash flow so low?
$554 million against about $3.9 billion of non-GAAP operating income, or conversion of about 14%. The main reasons are inventory build and investment in supply chain capacity.
What are AMD’s customer warrants and why do they matter?
They are part of strategic agreements with large AI customers. The change in their fair value (in the “Other” line together with corporate costs) reduces GAAP income but is excluded from non-GAAP. On exercise, they dilute shareholders.
What is the difference between $6 billion and $8 billion in Qualcomm Automotive?
$6.0 billion is reported revenue for full fiscal 2026 (+52%). About $8 billion is the run-rate - the record Q4 ($1.99 billion) multiplied by four. The latter shows the exit pace of the year but is not a reported annual result.
What does Qualcomm need to offset the handset decline?
Automotive and Data Center need to add about $3-4 billion a year. That is about 40-50% growth against a run-rate base and about 50-70% against reported FY2026 Automotive. For now, Automotive and IoT cover about 56% of the quarterly Handsets decline.
How do high memory and packaging prices affect margins?
Assuming they represent 25-35% of cost of goods, AMD’s Data Center gross margin could contract by 3-5 percentage points and overall non-GAAP gross margin by 2-3 points. At a share of about 40%, compression exceeds 5 points.
Which position is preferred? At the November 5, 2026 close: AMD $404.40 (-17.3%), Qualcomm $156.75 (-9.5%). AMD with full execution (conversion of 40% or more), now with a better relative entry after the session. Qualcomm in the middle scenario for AMD with successful diversification; trailing P/E about 12.3x. A lower price is not an automatic buy.
Not investment advice. This article is analytical material for educational purposes. Scenarios and P/E multiples are the author’s assumptions. Report data are for periods ended September 26 and 27, 2026. Market prices are as of the regular-session close on November 5, 2026 (Yahoo Finance).
