Home / Companies / AMD

Company research overview

AMD Research Dossier

Advanced Micro Devices, Inc. is researched through CPUs, GPUs, data center chips, and client computing. This dossier separates the business question, financial evidence, market context, and model observations.

Information TechnologySector baseline XLKReviewed 2026-09-02
Daily observations35historical sample
Distinct signal dates35not independent trades
5-day average0.57%next-open entry basis
5-day excess0.59%versus QQQ

Research brief

What has to be true

Separate end-demand from inventory restocking, then test whether mix and utilization support margins through the cycle.

This dossier is a structured starting point, not a current valuation or recommendation. Financial statements and market evidence must be refreshed before drawing a time-sensitive conclusion.

Business map

Translate the story into observable drivers

Operating focus

Cpus, gpus, data center chips, and client computing.

The useful question is whether reported economics support the narrative across a full cycle, not whether the narrative sounds plausible in one quarter.

Sector lens

Product cycles, platform durability, customer budgets, and operating leverage.

  • Enterprise and cloud budgets
  • Unit growth and pricing
  • Product transition timing
  • Research and capital intensity

Financial diagnostics

Five checks before a valuation conclusion

Measure to compareWhat to testEvidence trail
Organic Revenue GrowthTrack the level, direction, and management explanation across at least four reporting periods.Annual and quarterly filings; reconcile non-GAAP presentations to reported statements.
Gross MarginTrack the level, direction, and management explanation across at least four reporting periods.Annual and quarterly filings; reconcile non-GAAP presentations to reported statements.
R&D As A Share Of SalesTrack the level, direction, and management explanation across at least four reporting periods.Annual and quarterly filings; reconcile non-GAAP presentations to reported statements.
Free Cash Flow MarginTrack the level, direction, and management explanation across at least four reporting periods.Annual and quarterly filings; reconcile non-GAAP presentations to reported statements.
Remaining Performance Obligations Or BacklogTrack the level, direction, and management explanation across at least four reporting periods.Annual and quarterly filings; reconcile non-GAAP presentations to reported statements.

Model evidence

What the V4.67.1 replay observed

These are overlapping daily observations from V4.67.1 historical sample. They are not independent trades and do not include a claim of future performance.

AMD average forward return and excess return
Raw returnExcess vs QQQ
-1.8%0.7%3.3%5.8%8.4%1d3d5d10d20d

Thesis stress test

Define what would change the conclusion

Evidence strengthening

Core drivers improve while cash conversion, balance-sheet flexibility, and sector-relative performance remain consistent.

Evidence unresolved

The narrative improves but reported economics, guidance, or peer confirmation remain mixed.

Evidence weakening

Reported results contradict the driver map, management changes definitions, or the downside requires increasingly generous assumptions.

Risk register

Company questions that deserve a dated answer

Rapid Product Substitution

Define the first measurable evidence that this risk is moving from a possibility into reported results.

Customer Concentration

Define the first measurable evidence that this risk is moving from a possibility into reported results.

Inventory Corrections

Define the first measurable evidence that this risk is moving from a possibility into reported results.

Valuation Compression

Define the first measurable evidence that this risk is moving from a possibility into reported results.

Peer context

Nearby dossiers

Open the Information Technology sector board

Primary sources first

Evidence trail

Documents and market data change. Verify the latest filing and state the observation date before relying on a conclusion.

Model: V4.67.1 State ResearchEntry assumption: next-session openHorizons: 1/3/5/10/20 trading daysPage reviewed: 2026-09-02

Complete historical result table

AMD results across every measured horizon

The chart above is easier to scan, while this table supplies the values needed to check the interpretation. Returns use the next-session open assumption described in the data methodology. “Positive observations” is the share of overlapping observations with a positive raw close return; it is not a win rate from independent trades.

Forward horizonAverage raw returnExcess vs QQQExcess vs SPYPositive observationsAverage maximum upsideAverage maximum downside
1 trading day-0.59%-0.24%-0.44%54.3%2.31%-3.40%
3 trading days-0.83%-0.32%-0.84%45.7%5.46%-6.53%
5 trading days0.57%0.59%0.11%48.6%7.54%-8.16%
10 trading days4.05%3.19%2.56%65.7%11.45%-9.67%
20 trading days7.39%6.45%4.96%48.6%19.30%-13.12%
Sample reading: AMD had 35 overlapping observations on 35 distinct signal dates, grouped into 8 separated episodes. That is a limited sample. The strongest QQQ-relative average was 6.45% at 20 days; the weakest was -0.32% at 3 days. These describe this historical sample only and do not establish a forecast.

Company-specific research questions

What to verify for Advanced Micro Devices, Inc.

Advanced Micro Devices, Inc. is included because its operating story centers on CPUs, GPUs, data center chips, and client computing. A useful update should connect that story to reported figures, explain any definition change, and compare the same period with an appropriate peer or benchmark.

Cpus: demand

Locate the latest reported demand indicator for CPUs. Separate price, volume, mix, currency, acquisition, and timing effects where management discloses them. Record both the current period and the comparable prior period.

Gpus: economics

Test whether growth in GPUs is reaching gross profit, operating income, and cash flow. Note stock-based compensation, restructuring, capitalized costs, working-capital movements, or other items that make adjusted and reported outcomes diverge.

Data Center Chips: durability

Identify the earliest measurable evidence that data center chips is strengthening or weakening. Useful evidence can include retention, backlog, utilization, unit economics, concentration, regulation, or competitive pricing, depending on the company’s disclosures.

Quarterly update standard

Turn the next filing into a comparable update

Evidence layerMinimum information to recordReason it matters
Reported resultsPeriod, accounting basis, segment or consolidated scope, current value, prior-period value, and the exact filing page.Prevents a headline number from being compared with a differently defined figure.
Management explanationQuantified price, volume, mix, currency, acquisition, cost, or timing effects; label unquantified explanations separately.Shows which part of the change is measured and which part remains narrative.
Cash and balance sheetOperating cash flow, capital expenditure, material working-capital changes, debt, liquidity, and share count.Tests whether earnings quality and financing capacity support the operating story.
Information Technology comparisonUse the same dates and definitions for a peer or sector baseline; explain any unavoidable mismatch.Separates company-specific movement from a broad sector or market move.

A conclusion is ready for publication only after the source, period, calculation, counter-evidence, and next checkpoint are visible. If the filing does not disclose a needed figure, state that limitation rather than estimating it as fact.