A credible path exists to roughly double enterprise value by 2031 if platform leadership persists and revenue becomes more contracted and trust-driven. The swing factors are export licensing volatility and whether power constraints and custom silicon meaningfully slow demand or compress pricing.

In the Last Economy, “compute-to-outcome” becomes core infrastructure: NVIDIA can keep compounding by staying the default AI-factory platform (accelerators + networking + systems + ops software), and by shifting more revenue into durable, higher-trust layers (verified infrastructure, optimization and contracted capacity) even as export controls and power availability throttle the slope.

NVIDIA sits on the compute-supremacy flywheel: it turns capital into scarce, high-leverage compute with ecosystem lock-in (developer workflows + OEM/cloud distribution), and can extend into trust/verification as AI offense outpaces defense.

NVIDIA’s edge is time-to-deploy and time-to-performance: buyers optimizing for fastest usable AI output tend to prefer an integrated platform (accelerators + networking + systems + software) over point chips. Over 5 years, the upside is less about “more GPUs” and more about higher content per deployed megawatt (systems/networking attach), broader buyer mix (enterprise + national deployments), and more recurring monetization (contracted capacity, verification/security, and performance optimization). That durability helps defend a premium revenue multiple even as growth normalizes from peak levels.

The two binding risks are exogenous: (1) export-control volatility that can shrink addressable demand and force product re-segmentation, and (2) power/grid constraints that slow customer deployment regardless of NVIDIA demand. Endogenously, the key risk is value capture shifting toward custom silicon and commoditized inference, which would pressure pricing power and compress the premium platform multiple even with continued unit growth.

No material change: platform leadership remains central, but export licensing has become a more explicit near-term gating variable for reachable demand and volatility.

Slightly more conservative on the re-rating path due to clearer policy gating and the probability of gradual multiple compression with scale.

$253.62