Friday, September 11, 2026

Watch —Primary condition unchanged: Nvidia Q3 FY2027 REPORTED actual revenue and gross margin — quarter ending approximately October 26, 2026, results expected late October or early November 2026; REPORTED actuals only, not the $108 billion Q3 guidance already in hand.

If Q3 reported revenue is at or above $108 billion with gross margin at or above 74% and no explicit demand-deferral or pricing-pressure language from hyperscaler customers on the earnings call, hold Fragility at Present and the phase at Structural Expansion.

If Q3 reported gross margin declines below 73%, that constitutes the first verified trigger for escalating Fragility from Present to Acute — this branch requires an affirmative adverse finding on the actual disclosed figure, not silence or a figure simply not restated.

If explicit deferral or pricing-pressure commentary from hyperscaler customers appears on the Nvidia Q3 earnings call, that is a second independent trigger for the same Fragility escalation — the two branches are distinct and either one is sufficient.

Google TPU InferenceX: monitor for any named enterprise or hyperscaler customer publicly disclosing that it is shifting inference workloads from Nvidia CUDA to TPU InferenceX at material scale, or any Nvidia earnings-call language acknowledging inference pricing pressure from internal-silicon alternatives — either would confirm this signal has moved from competitive dynamic to structural demand displacement.

Speculation Pressure escalation from Moderate — Elevated to Elevated requires: a second named hyperscaler (beyond Google) publicly disclosing a commercial inference product or confirmed workload redirection away from Nvidia CUDA, OR Nvidia explicitly acknowledging inference pricing pressure from internal-silicon alternatives on an earnings call — neither condition is met as of this cycle.

DeepSeek hardware independence escalation: monitor for any Western enterprise or government cloud provider publicly adopting or evaluating the Huawei-based DeepSeek stack for production workloads — if confirmed, that would represent a new Narrative Drift escalation input independent of Nvidia or Broadcom earnings.

Broadcom Q4 FY2026 REPORTED actual revenue — quarter ending approximately November 2, 2026, results expected early December 2026; the $34.8 billion Q4 guidance is already in hand and does not satisfy this condition; monitor for any hyperscaler commentary on rising memory costs accompanying the reported actuals.

HBM pricing: any confirmed announcement of HBM contract price increases taking effect in 2027 that hyperscaler customers explicitly acknowledge in earnings commentary would constitute an emerging cost-pressure signal worth incorporating into the Fragility dimension, independent of Nvidia or Broadcom earnings timing.

NSA-FBI-CISA advisory watchpoint: if any Western government agency formally restricts enterprise or government use of the named Chinese AI models (DeepSeek, Moonshot AI, Alibaba) in procurement or national-security contexts, that represents a confirmed regulatory escalation worth incorporating into the Narrative Drift dimension.

California AI-auditor law: monitor for any named enterprise publicly citing compliance cost or operational restriction under the newly signed California law as a reason to defer or condition AI infrastructure spending — if this pattern appears in procurement language, it would constitute a Narrative Drift escalation input.

BIS Affiliate Rule resumption on November 10, 2026 remains a hard operational deadline — any confirmed enforcement action or material operator withdrawal tied to that date before it arrives should be noted as an emerging fragility escalation signal independent of earnings.