AMD · one lens
Hamilton Helmer on AMD
AMD shows demand and cash-flow growth, but the record has not proved a barrier that turns either into continuing Power.
AMD’s record clears Helmer’s Significant test more readily than Sustainable. Per the TEXXR company dossier, AMD’s fiscal 2025 revenue rose 34%, to $34.64 billion from $25.79 billion in fiscal 2024, while free cash flow rose 180%, to $6.74 billion from $2.41 billion. The market record shows AMD rose 4.69%, to $477.57 on September 4 from the prior $456.16 close. The record does not turn either result into Power: Helmer requires a cash-flow Benefit and a named Barrier against arbitrage. The record shows cash-flow improvement but does not isolate a Superior differential versus rivals.
AMD has not yet demonstrated Scale Economies. AMD has signed large Meta and Anthropic accelerator commitments and introduced the Helios rack-scale system, which supports a volume opportunity. The record supplies no unit-cost curve, no leader-versus-challenger scale gap, and no evidence that AMD can price where a smaller rival loses money while AMD remains profitable. AMD’s higher cash flow is Significant, but the record does not tie it to that mechanism.
OpenAI and Broadcom developed Jalapeño from project start to chip in 16 months, and SemiAnalysis reported that the ASIC beat AMD, Nvidia and Google chips on multiple top open-weight models. Jalapeño pressures Sustainability because OpenAI built an alternative to merchant accelerators rather than merely choosing among them. The report does not prove broad superiority across workloads, but it weakens any claim that AMD has a non-replicable technical resource.
Mercury Research’s client CPU share record says AMD accounted for 30.3 of every 100 shipments in the second quarter of 2026, versus 21.1 of every 100 in the second quarter of 2024; Intel still accounted for 69.7 of every 100 in the later period. AMD therefore shows share gains, not the leadership gap Helmer’s Scale Economies test needs.
AMD may deepen customer integration through Helios, but the record shows no separate migration cost that makes customers stay. AMD acquired Taalas, whose design integrates model weights into silicon, but the record does not establish preferential access that rivals cannot obtain at any price. The September 5 Zenith item describes Microsoft’s developer experience and establishes no AMD-specific cost, price, or lock-in benefit.
The TEXXR record makes no Network Economies, Counter-Positioning, Branding, or Process Power case. The record also shows competitor predicates at 1.2%, or six of its 500 most recent tracked AMD edges, a description of coverage extraction rather than real-world rivalry. AMD has commercial validation; the record has not yet established continuing Power.
The seven Powers, one by one
| Power | Status | Why |
|---|---|---|
| Scale Economies | contested | revenue and customer commitments support volume, but no unit-cost curve or leadership gap establishes the Barrier |
| Network Economies | absent | not argued in the record |
| Counter Positioning | absent | not argued in the record |
| Switching Costs | contested | Helios adds rack-level integration, but the record shows no separate customer cost to leave |
| Branding | absent | not argued in the record |
| Cornered Resource | contested | Taalas adds owned technology, but the record does not show an asset unavailable to rivals at any price |
| Process Power | absent | not argued in the record |
An editorial application of Hamilton Helmer's published framework — not their actual view, affiliation, or advice.
Where this lens comes from
Hamilton Helmer's framework is set out at the lens page, drawn from 7 Powers. This page applies it to AMD and nothing else — the company's full record is in the dossier.