NVDA is resilient over 2,028d of continuous history. NVIDIA is the clearest public-market AI infrastructure winner here. Core bars 3/3: post-maturity PayBack 11.37% clears 8.00%; seasoned PayBack 14.25% clears 8.00%; B/D 9.10x clears 1.20x. PayBack billing: 8,989 events, $193.3M due. OFL gate open 100.0%; throughput 993%. Temporal-hedge timing: $156.5M of PayBack remains deferred (80.9% of cumulative due); ARC records the obligation and pays retrospectively from earned inflows only. Stress load: 3 modeled bear-panic episodes, 1 incident, $250.0M hack absorption. Simulated 2026.07.22.0822.
NVDA ran a powerful long-term uptrend: end-to-end +1,486.0% over 2,028 sessions after a peak +1,703.7% above the open, while still absorbing a 66.4% peak-to-trough decline and 2 multi-month drawdown regimes. It closed 12.1% off the window high; this is not a completed boom-and-bust cycle. Against that tape, system throughput finished 993%; Class A reads should emphasize buffer defense and PayBack service more than matching every point of asset upside.
Cumulative floor payments and Daily PayBack share the purple PayBack color; asset price overlay is cyan. Zone pane: Supercharge top half, Strategic Growth bottom half — tier depth by opacity.
Buffer / Defense
Two independent scales: B/D Ratio measures whole-buffer defense against the worst observed 30-day withdrawal principal requested plus its attributable vested OFL; FCR measures short-horizon floor cadence against current OFL. Purple is the reported B/D 365-day SMA; green is FCR. The B/D floor is 1.20x.
Milestone funding, cyclic-buffer movements, and SC/SG zones use a synchronized time scale across panes.
NVIDIA is the clearest public-market AI infrastructure winner here. The integration case is extraordinary cash generation and CUDA lock-in; the risk is how much of the thesis depends on hyperscaler capex and power delivery continuing without a sharp air pocket.
Revenue is concentrated in accelerated computing sold into cloud and sovereign buildouts. That concentration is the moat (when demand is hot) and the fragility (when customers pause purchases after overbuild).
Full-stack software and networking around GPUs raise switching costs far above raw silicon. Competitors can win sockets, but displacing the developer ecosystem remains hard — a structural support for multi-year stress tests.
CUDA’s installed base is itself a network effect. For ARC, that supports resilience of the equity franchise, while valuation still prices near-perfect execution.
Cash generation is exceptional in boom phases. Underwrite paths for what happens when growth decelerates from extreme rates: buffer defense must survive multiple compression, not only chip shortages.
Public debate is no longer “will AI spend happen?” but “who pays for power, and when does ROI show?” Sentiment is positive with sharper questions on duration of the spend cycle.
Observation window: mid–late July 2026. NVIDIA social narrative remains the AI infrastructure supercycle versus power, export-control, and customer-capex payback debates. Bulls anchor on hyperscaler demand and CUDA lock-in; bears fixate on valuation and whether inference economics normalize margins. For ARC, equity-class stress is growth-multiple compression under capex disappointment—not product irrelevance.
Public Inspect is the discussion summary. The full interactive canvas (every series, flashcards, parameters, and internal notes) lives under the gated Internal reports area.
