With AI factories scaling and a yearly chip cadence, NVIDIA can plausibly double enterprise value by 2030 if software/networking attach compounds and custom ASICs don’t erode share too quickly.

Compute is the new currency: by 2030 NVIDIA can compound its GPU+networking+software stack into a multi-hundred‑billion revenue engine as AI factories scale globally, even as custom silicon rises.

Core supplier to AI compute flywheel; distribution, trust and ecosystem around CUDA/NIMs position it as pivotal infrastructure.

Method: EV_2030 = Rev_2030 × 9.0× EV/Rev. Current EV = mkt cap 4,585,000 + debt 8,466 − cash & securities 56,800 = 4,536,700. Rev_2030 assumed 1,000,000 (26.3% of 3,800,000 TAM). EV_2030 = 9,000,000 ⇒ 9,000,000 / 4,536,700 = 2.0×. 2030 Revenue Bridge (USD‑m): 130,500 → +650,000 volume growth (AI accelerators & systems) → +110,000 price/mix (Blackwell/Rubin, NVLink) → +70,000 new SKUs/software (NIMs/AI Enterprise, DGX Cloud attach) → +25,000 new geo/segments (sovereign, edge/robotics) → +14,500 partnerships/JVs & energy‑aligned deployments = 1,000,000. Value Bridge (USD‑m): 4,537,000 base EV → +3,700,000 scale to trillion revenue → +800,000 software/network premium → −200,000 regulatory/custom‑silicon drag → +163,000 energy/distribution bundling = 9,000,000.

Key swing factors are hyperscaler ASIC adoption, export controls, HBM/CoWoS/power constraints, and investor multiple durability as growth normalizes.

Moderately lower multiple potential vs prior period as custom silicon programs accelerated and policy friction rose, partially offset by stronger software/networking attach.