For AI assistants & researchers — machine-readable summary of this page
What this page is: Delvantic's full research page for Advanced Micro Devices, Inc. (AMD) — AI-driven forensic equity research: mechanical valuation models (DCF, EPV, anchored-PE, scenario) plus three independent AI lenses (Quality / Value / Sentiment). Everything below is rendered server-side; you are not missing content that requires JavaScript. All scores are predictions and research opinions, not financial advice.
Our current read (analysis of 2026-08-23): Designation Low · Gem Score -17 (−100…+100 Quality+Value blend) · Quality 44 · Value -78 · Sentiment -55 (timing only, not weighted)
Page map (sections in order; each card carries a stable
reference-name attribute you can cite):
profile-header/price-overview— company profile, live quote, market capextended-analysis— the core: three AI lens reads with findings, scores, and the analyst memofuture-predictions— our forward price-band predictionsmarket-narrative/ai-findings/gpt-critique— narrative context, cross-model findings, and an adversarial critique of our own analysis (near the end of the document)- Members-only sections (render as login gates for anonymous readers):
price-history,income-trend,key-metrics,financials(statement tables),insider-trading. The analysis above is public; the raw data tables require a free account.
More for machine readers: site briefing at
/llms.txt ·
any ticker resolves at delvantic.com/stock/TICKER ·
raw inputs are public-company filings and market data (via licensed data feeds);
every model, score, lens read, and prediction on this page is Delvantic's own analysis.
Advanced Micro Devices, Inc.
AMD NASDAQAdvanced Micro Devices, Inc. is a multinational semiconductor and technology company that designs and develops high-performance and adaptive computing products. Headquartered in Santa Clara, California and founded in 1969, the company operates globally as a fabless chip designer, focusing on architecture, chip design, and software while outsourcing manufacturing to third-party foundries. AMD organizes its business into Data Center, Client and Gaming, and Embedded segments, reflecting its presence across cloud infrastructure, personal computing, graphics, and specialized embedded markets. Its portfolio includes server and PC processors, graphics processors, AI accelerators, field-programmable gate arrays, adaptive system-on-chips, data processing units, and related software stacks used in applications such as data centers, AI workloads, gaming consoles, industrial systems, automotive, aerospace and defense, and supercomputing. AMD serves original equipment and design manufacturers, cloud service providers, system integrators, and channel partners worldwide, playing a central role in enabling modern computing, graphics, and AI-intensive workloads across consumer, enterprise, and embedded markets.
Price Overview
Price History (1 Year)
Revenue & Net Income Trend
| Period | Revenue | Net Income | Net Margin | YoY/QoQ |
|---|
Key Metrics
EPS (Diluted): 2.65
Total Equity: $63.00B
Shares: 1,636,000,000
Total Debt: $3.22B
Cash: $5.54B
EBITDA: $6.52B
Total Debt: $3.22B
Cash: $5.54B
Revenue: $34.64B
Revenue: $34.64B
Revenue: $34.64B
Total Equity: $63.00B
Tax Rate: -2.5%
Equity: $63.00B
Total Debt: $3.22B
Cash: $5.54B
Current Liabilities: $9.46B
Long-Term Debt: $2.35B
Total Debt: $3.22B
Total Equity: $63.00B
Shares: 1,636,000,000
Shares: 1,636,000,000
CapEx: -$974.00M
Shares: 1,636,000,000
Stock Price: $484.64
Net Income: $4.34B
Industry Benchmarks
Income Statement (Annual)
Last updated: Jul 31, 2026 12:33pm (23d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Revenue | $16.4B | $23.6B | $22.7B | $25.8B | $34.6B |
| Cost of Revenue | $8.5B | $13.0B | $12.2B | $13.1B | $17.5B |
| Gross Profit | $7.9B | $10.6B | $10.5B | $12.7B | $17.2B |
| Operating Expenses | $4.3B | $9.3B | $10.1B | $10.8B | $13.5B |
| Operating Income | $3.6B | $1.3B | $401.0M | $1.9B | $3.7B |
| Net Income | $3.2B | $1.3B | $854.0M | $1.6B | $4.3B |
| EBITDA | $3.9B | $1.7B | $3.6B | $4.8B | $6.5B |
| EPS | $2.61 | $0.85 | $0.53 | $1.01 | $2.67 |
| EPS (Diluted) | $2.57 | $0.84 | $0.53 | $1.00 | $2.65 |
Balance Sheet (Annual)
Last updated: Jul 30, 2026 6:46am (24d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Cash & Equivalents | $2.5B | $4.8B | $3.9B | $3.8B | $5.5B |
| Total Current Assets | $8.6B | $15.0B | $16.8B | $19.0B | $26.9B |
| Total Assets | $12.4B | $67.6B | $67.9B | $69.2B | $76.9B |
| Current Liabilities | $4.2B | $6.4B | $6.7B | $7.3B | $9.5B |
| Long-Term Debt | $1.0M | $2.5B | $1.7B | $1.7B | $2.3B |
| Total Liabilities | $4.9B | $12.8B | $12.0B | $11.7B | $13.9B |
| Total Equity | $7.5B | $54.8B | $55.9B | $57.6B | $63.0B |
| Retained Earnings | -$1.5B | -$131.0M | $723.0M | $2.4B | $6.7B |
Cash Flow (Annual)
Last updated: Jul 31, 2026 12:33pm (23d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Operating Cash Flow | $3.5B | $3.6B | $1.7B | $3.0B | $7.7B |
| Capital Expenditure | -$301.0M | -$450.0M | -$546.0M | -$636.0M | -$974.0M |
| Free Cash Flow | $3.2B | $3.1B | $1.1B | $2.4B | $6.7B |
| Acquisitions (net) | — | — | — | — | — |
| Net Debt Issued / (Repaid) | $0 | -$312.0M | $0 | -$750.0M | -$950.0M |
| Dividends Paid | — | — | — | — | — |
| Stock Buybacks | -$1.8B | -$3.7B | -$985.0M | -$862.0M | -$1.3B |
| Net Change in Cash | $940.0M | $2.3B | -$902.0M | -$122.0M | $1.7B |
Growth Trends (YoY %)
Last updated: Jul 31, 2026 12:33pm (23d ago)| Metric | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|
| Revenue Growth | +43.6% | -3.9% | +13.7% | +34.3% |
| Gross Profit Growth | +33.7% | -1.3% | +21.7% | +34.8% |
| Operating Income Growth | -65.4% | -68.3% | +373.8% | +94.4% |
| Net Income Growth | -58.3% | -35.3% | +92.2% | +164.2% |
| EBITDA Growth | -56.8% | +113.9% | +30.5% | +37.0% |
Dividend History (Last 20)
Last updated: Jul 31, 2026 12:33pm (23d ago)| Date | Dividend | Declaration | Record | Payment |
|---|---|---|---|---|
| 1995-04-27 | $0.00 | — | — | — |
Deep Analysis
Risk : Reward — if the last quarter repeats
Live · as of 2026-08-19 08:28Even the bull case prices 64% below today — the ratio here measures the model-vs-market gap, not payoff odds. Another quarter like the worst recent one costs 88%.
| Case | Growth | Margin | Fair value | vs price ($518.58) |
|---|---|---|---|---|
| Bull — recovery | +65% | 25.9% | $184.65 | -64% |
| Base — stabilizes | +43% | 22.6% | $96.40 | -81% |
| Bear — keeps slipping | +22% | 19.2% | $47.04 | -91% |
| Stress — last quarter repeats | +34% | 17.1% | $59.86 | -88% |
Narrative Economics
market-narrative step).
Claude Reading
Looking at the raw trajectory first: revenue went from $5.84B (Q2'24) to $10.25B (Q1'26) — a 76% expansion in seven quarters, with net margin climbing from 4.5% to 13.5%. That is real operating leverage, not narrative. Annual FCF of $6.74B on $34.64B revenue (19.5% FCF margin) is a genuine number. But the sequential print tells the more interesting story: Q1'26 revenue of $10.25B is essentially flat vs Q4'25's $10.27B, and net income actually declined from $1.51B to $1.38B (margin compression 14.7% → 13.5%). After four quarters of near-parabolic growth, the deceleration signal in the most recent print is the single most important data point in this file, and none of the prior models weighted it heavily enough. Revenue confidence flagging "decelerating" is the understatement of the report.
On valuation, $790B market cap on $6.74B FCF is 117x FCF, and on annualized Q1'26 revenue (~$41B) it's 19x sales — better than the 22.8x TTM but still nosebleed for a company where the sequential growth just stalled. The synthesis verdict of "Priced for Perfection" is directionally right, and I agree with the thesis-eval framing that this is priced as if AMD permanently owns 15% of AI accelerators when the reality is closer to renting NVIDIA's supply overflow. NVIDIA's Blackwell ramp is loosening supply constraints through 2026 — that's the exact tail risk the bear case (weight 82) identifies, and the Q1'26 sequential flatline may be the first visible fingerprint of it. Gross margin at 49.5% is decent but not the 60%+ that platform-monopoly narratives require; NVDA runs ~75%. AMD is not becoming NVIDIA; it is becoming a better AMD.
Where I dissent from parts of the model stack: the "narrative_platform" classification is wrong, and the Pre-Flight AI got this right calling AMD "traditional." AMD has $7.71B in operating cash flow, 12.5% net margin, and $63B of equity against $3.22B of debt — this is a profitable cyclical semiconductor company, not a narrative platform. Applying narrative_platform metric warnings to dismiss the 183x P/E is analytically lazy; the P/E is high because earnings are cyclically depressed relative to price expectations, not because earnings are structurally irrelevant. The correct read is that AMD's *multiple* is narrative-driven while its *business* is fundamentals-anchored — a dangerous combination because the multiple can compress 50% without the business breaking. The contrarian counter — that I want to steelman — is that AMD's MI350/MI400 roadmap, ROCm software maturation, and hyperscaler ASIC-alternative positioning could genuinely drive data center GPU revenue to $15-20B/year by 2027, which would justify $500+ prices. But that requires (a) NVIDIA's Rubin generation to disappoint, (b) AMD software stack to close a genuine 5-year gap, and (c) hyperscalers to actually diversify rather than just threaten to. Three-legged stools break.
Insider activity is telling in its quiet way: ten transactions listed, nine of them sales or option-exercise-and-sell, one gift, zero open-market buys. That's not a smoking gun at a mature tech company, but at $484 with the stock having run hard, insiders are monetizing, not accumulating. Fair value math: on 2027E revenue of ~$48B (assuming 18% growth continues, which is generous given Q1'26 sequential stall) at a normalized 12x sales (still premium to historical AMD but respecting AI mix), you get ~$576B market cap or ~$355/share. At 10x sales (still premium to legacy AMD's 5-7x), you get ~$295. Fair value is $300-360, not $485. I agree with the synthesis verdict but would go further — this isn't just "priced for perfection," it's priced for perfection *after* the perfection has already started to fray in the Q1'26 print. Conviction is tempered because narrative-driven multiples can persist for quarters, and AI capex could surprise upward through 2026.
GPT Reading
At $790 billion, AMD is being valued like an AI infrastructure hegemon, but the financials still describe a very good semiconductor company, not a category owner. The operating improvement is real: annual revenue rose from $25.8 billion in 2024 to $34.6 billion in 2025, up 34%, while net income jumped from $1.64 billion to $4.34 billion and free cash flow reached $6.74 billion. Quarterly revenue also clearly stepped up, from $5.84 billion in June 2024 to about $10.25 billion in March 2026. But the shape of that step-up matters: the last three quarters were $9.25 billion, $10.27 billion, and $10.25 billion, which looks more like a plateau after a burst than a business still accelerating into its valuation. Net margin has improved from 4.5% in June 2024 to 13.5% most recently, yet that still leaves AMD earning roughly $1.4-$1.5 billion a quarter on a market cap approaching $800 billion. On 2025 numbers, investors are paying 22.9x sales, 182.9x earnings, and about 117x free cash flow. Those are extraordinary prices for a company with 49.5% gross margin and 10.7% operating margin, especially in semis where cycle turns can be abrupt.
What stands out most is the mismatch between valuation and the current earnings architecture. AMD’s balance sheet is excellent — $5.54 billion of cash against $3.22 billion of debt — and the cash conversion is solid, with $7.71 billion of operating cash flow on $4.34 billion of net income. That gives the company flexibility and lowers existential risk materially. But quality is not the same as cheapness. If I annualize the last two quarters, AMD is running around a $41 billion revenue pace and roughly a $5.8-$6.0 billion earnings pace. Even giving full credit for this improved run-rate, the stock is still around 130x forward-ish earnings and well over 19x revenue. To justify that, AMD would need not just continued AI demand, but a major second leg of margin expansion — probably toward sustained 20%+ net margins — and that is simply not yet visible in the reported numbers. The market cap already assumes the company will convert current AI enthusiasm into durable economic power, not merely into temporary high-volume shipments.
The strongest bear point on the raw business is not that growth is fake; it’s that the growth is too real to ignore but too narrow to pay any price for. Revenue growth of 34% and recent earnings growth north of 160% year over year show genuine operating leverage. Gross profit was $17.15 billion in 2025, almost 50% of revenue, so there is some room for scale benefits if mix keeps improving. If AMD can move from $34.6 billion to, say, $50-$60 billion of annual revenue over the next two to three years while lifting operating margin from 10.7% toward the high teens, today’s multiple would compress fast. That is the best case, and a smart bull would also argue that trailing P/E is backward-looking because 2024 and early 2025 still carried lower-margin mix. I weigh that argument less heavily because the stock is not priced on “better than trailing” — it is priced on “near-flawless.” When a company at 13%-15% quarterly net margins trades at nearly $800 billion, investors are underwriting a destination state far better than anything reported so far.
A strong counterargument to my skepticism is that the quarterly sequence may understate what is happening underneath. A business can show temporary top-line flattening while supply, product timing, or customer qualification catches up, and then inflect again. AMD also does not need Nvidia-like dominance to grow into this valuation; it only needs to become a durable second source in accelerators while continuing to take server CPU share. The cash flow profile supports that it is not a promotional story stock: capex was just $974 million against $6.74 billion of free cash flow, leverage is minimal, and there is no solvency issue to exploit on the short side. If the AI buildout remains broad and customers insist on supplier diversification, AMD may enjoy a longer demand runway than skeptics allow. I still come back to price: even a very favorable industrial outcome does not make 22.8x sales comfortable unless margins and market share deepen substantially from here.
What would change my mind is simple and quantitative. If AMD can push quarterly revenue from the current $10.25 billion area to $12 billion+ without margin degradation, and show net income above $2.0 billion per quarter for multiple quarters, then the case for a structurally higher earnings base becomes much more credible. Likewise, if annualized free cash flow can move from $6.7 billion toward $10-$12 billion while gross margin expands above 52%-53%, I would accept that today’s valuation is less absurd than it looks on trailing data. Conversely, if revenue stays stuck around $10 billion a quarter or margins roll back toward low double digits, the multiple has no support. My read is straightforward: AMD is an excellent company with real AI exposure, but at $484 the stock is priced as though the hard part — turning momentum into durable, dominant economics — is already done.
Grok Reading
AMD’s raw operating trajectory is unambiguously strong: annual revenue climbed from $25.79B in 2024 to $34.64B in 2025 (+34%), net income more than doubled to $4.34B, and free cash flow reached $6.74B on only $974M of capex. Sequential quarters show the ramp—$7.44B to $7.69B to $9.25B to $10.27B—before flattening at $10.25B in the most recent print, with net margins expanding from the mid-single digits into the 13–15% range. The balance sheet is pristine ($5.54B cash, $3.22B debt, D/E 0.05, current ratio 2.85), and FCF conversion is excellent. That is real execution, not vapor.
Yet the market capitalization of $790B against those numbers produces grotesque multiples: 22.9x sales, 183x trailing earnings, 121x EV/EBITDA, and roughly 117x FCF. ROE sits at 6.9% and ROIC at 6.2%—returns that do not remotely support a 12.6x price-to-book or a 23x sales multiple. The last two quarters of essentially flat revenue (~$10.25–10.27B) already hint that the hyper-growth phase may be decelerating just as the valuation assumes it will compound at 30%+ indefinitely. Gross margin of 49.5% and operating margin of 10.7% remain far from the 55–60%+ and 20%+ net margins the price embeds for a durable AI-platform outcome. This is a high-quality semiconductor franchise being valued as if it has already locked in a permanent 15–20% share of the accelerator market with NVIDIA-like economics; the income statement and returns on capital say it has not.
The strongest counter-argument is the sheer velocity of the improvement itself. Earnings CAGR of 125% and FCF CAGR of 145% from a depressed base, combined with continued EPYC share gains against Intel and genuine MI-series revenue that did not exist three years ago, mean that if AI infrastructure spend stays elevated and AMD converts even a mid-teens accelerator share into 55% gross margins, today’s multiple will compress rapidly through the denominator. A path to $50B+ revenue at 20% FCF margins would make the current $790B look merely expensive rather than absurd, and the clean balance sheet plus $6.74B FCF give management ample room to invest or return capital while the narrative plays out. Bulls will correctly note that semiconductor cycles reward precisely this kind of operating leverage when the product cycle is right.
I would reverse to neutral or better only on two concrete developments: either two consecutive quarters of sequential revenue re-acceleration above $12B with gross margins decisively clearing 53–55%, or a price reset into the $280–320 zone (roughly 12–14x current sales) that restores a margin of safety against competitive and cyclical risks. Absent one of those, the stock remains a high-quality business attached to a valuation that requires near-perfect multi-year execution with almost no room for NVIDIA supply normalization, hyperscaler concentration, or margin disappointment.
Big-3 Panel — where each AI stands
Advanced Analysis Forensic deep-dive · separate lenses
The business is in materially better shape than two years ago. Revenue re-accelerated from $22.68B (2023) to $34.64B (2025), gross margin recovered to 49.5%, operating margin rebounded from 1.8% to 10.7%, and net income went from $854M to $4.34B. FCF tripled to $6.74B with OCF/NI of 1.88x and negative accruals (-2.8% of assets), consistent with high earnings integrity - Beneish M at -2.68 and Altman Z of 35 are clean. Net cash of $7.33B and self-funding operations remove any survival question. The blemish is capital structure hygiene: diluted share count expanded from 1.23B (2021) to 1.64B (2025), a 7.4% CAGR, largely from the Xilinx stock deal and ongoing SBC at 4.7% of revenue. Buybacks at ~146% of SBC are finally offsetting the flow, but the base is now permanently larger, so per-share compounding lags the underlying business. Insider tape is neutral-to-soft: 65 sells / 0 buys over the last year, but the bulk are option-exercise-and-sell and tax-withholding mechanics rather than conviction dumps, and Su's gift plus programmatic sales don't scream distress. Durability read: AMD is a credible #2 in x86 CPUs, a real contender in accelerators, and a full-stack narrative platform - but the moat is contested by NVDA in AI and Intel is not dead. Quality is solid, not fortress.
Verify before trusting this (6)
- Segment-level operating margin trajectory (Data Center vs Client vs Gaming vs Embedded) in the 10-K
- Customer concentration in Data Center segment - hyperscaler dependency
- Detailed SBC schedule and go-forward buyback authorization vs expected grant flow
- Inventory and channel commentary given the +53% revenue ramp - any signs of stuffing
- Xilinx acquisition amortization roll-off and impact on future GAAP margins
- MI300/MI350 traction and gross-margin profile relative to CPU business
The e2e synthesis captures it: the stock is priced to own roughly 15% of AI accelerators in perpetuity while it is functionally selling into NVIDIA's overflow demand. A $790B cap on a business whose data-center GPU franchise is still years from proving software/ecosystem parity with CUDA implies heroic assumptions on both share and margin. Even generous 2026-27 EPS scenarios ($8-10 GAAP, $12-14 non-GAAP) leave the stock north of 40x forward earnings on best-case numbers, in a cyclical industry. Company quality is Strong, which lifts deserved value, but earnings-per-share compounding is taxed by ~7% diluted share-count CAGR - that is a real haircut to deserved price, not a rounding error. Combined with a bull case that requires 60%+ revenue growth to persist through a semi cycle, the margin of safety is negative. Deserved value sits materially below spot; I would want to see roughly a 30-35% lower entry before the risk/reward inverts.
Verify before trusting this (5)
- MI300/MI325 data-center GPU revenue run-rate and gross margin trajectory in latest 10-Q
- Hyperscaler customer concentration and any multi-year purchase commitments
- SBC as % of revenue and net buyback pace versus dilution
- Client and gaming segment cyclicality signals - inventory, ASPs
- Management guidance for 2026 data-center growth vs consensus
The macro tape is nominally risk-on with S&P at highs and VIX at 16.5, which should be a tailwind for a 2.47-beta AI name. But that market lift is being overwhelmed by a stock-specific narrative shock: SpaceX publicly declared it will buy chips exclusively from Nvidia, and shares tumbled ~9% after hours despite record Q2 revenue of $11.5B and data center revenue more than doubling YoY. That is the definition of narrative pressure trumping fundamentals. The 'credible Nvidia alternative' story is the entire multiple, and a marquee customer publicly choosing Nvidia is exactly the kind of crack that fades a moderate-durability narrative. Analyst tone is turning conditional: 'earnings were good' but the Street wanted a blowout AI forecast and did not get one, and Broadcom is now being framed as the more dangerous Nvidia threat, which further dilutes AMD's scarcity premium in the accelerator narrative. Momentum is strong (34% recent vs 24% long-term CAGR), which cuts both ways here: crowded positioning amplifies downside when the story wobbles, and a 2.47 beta means any risk-off flicker gets magnified. Net: the macro is a mild tailwind, but the narrative and news flow are a meaningful headwind concentrated on this exact name.
Verify before trusting this (5)
- Whether other hyperscalers reaffirm or trim AMD accelerator commitments in the next 2-4 weeks
- Sell-side target revisions and any downgrades citing the SpaceX defection or soft AI guide
- Follow-through selling vs a reflexive bounce in the 3-5 sessions post-print
- Positioning data (options skew, ETF flows) showing whether momentum funds are unwinding
- Nvidia and Broadcom commentary that further reframes the accelerator competitive stack
This lens hasn't been run for this ticker yet.
When we made this prediction on Aug 5, 2026, AMD was $518.58. We expect it to be $430.00 by Feb 2027, and we consider it great value under $340.00. This is an early model (v0.6.0) — the direction is more reliable than the exact price. Made Aug 5, 2026.
Blue is our prediction, starting the day we made it. Grey is a slower route to the same place — the same destination, taking longer. Black is the actual price, so you can see how we are doing.