brain/
← all entities
entityNVDAstock-market

Nvidia

Notes

Nvidia

One-line summary: Dominant AI-accelerator (GPU) company; "probably the greatest company in the first part of this century" per a competitor (andrew-feldman). Tracked here mainly for the durability of its moat at inference — Feldman argues CUDA is "not important now and has no role whatsoever in inference," and that two of the three leading frontier models already train without CUDA.

What it is

Designs the GPUs (H100/H200/B200/GB300) and the CUDA software stack that underpinned the first wave of AI training. The GPU architecture is, per Feldman, "an extremely good architecture and extremely efficient at building very slow tokens" — i.e. cheap at low speed, but its cost and power-per-token rise steeply as you push for fast tokens. Nvidia is the single largest consumer of TSMC CoWoS capacity (>50%) and the anchor demand behind the HBM shortage.

Why it matters to artificial-intelligence

For the AI thread (the "compute markets / GPU supply" subdomain), the durable claim from this source is the decay of CUDA as an ecosystem moat, framed as a fact about where frontier labs actually run, not about NVDA's price. Per andrew-feldman, CUDA "was really important in the creating of the AI landscape" but is now irrelevant at inference and shrinking at training: of the three leading frontier models, Gemini (Google TPUs) and Anthropic's Claude (Trainium) use no CUDA, while only OpenAI's GPT trains in the CUDA environment. The architectural why — GPUs being optimized for cheap slow tokens, with cost/power-per-token rising as you push for speed — is the hardware constraint that the thread's inference-economics sources (Turley's gpu-as-zero-sum-constraint, the chatgpt-super-assistant-vision pricing forecast) discuss from the demand side. See cuda-moat-erosion-at-inference and open-vs-closed-source-model-economics.

Why it matters to stock-market

Nvidia is the incumbent every alternative-silicon thesis is implicitly short. Two SCOPE-relevant claims from this source:

  1. CUDA-moat erosion at inference. Feldman: CUDA was decisive in creating the AI landscape but "has no role whatsoever in inference," and a model can be moved from GPUs to Cerebras "in 10 keystrokes." A year ago every frontier model was CUDA-built; today two of three are not (Gemini on TPUs, Anthropic on Trainium) — "a hemorrhaging of share." See cuda-moat-erosion-at-inference.
  2. Export-control posture. Feldman explicitly comes down against Nvidia's stated position (give China access to keep them on US-designed product); he favors limiting diffusion of "our most precious technologies" to an "industrial enemy," accepting that some markets get foreclosed.
  3. Valuation / catch-up-trade view (dan-loeb, Third Point, May 2026). Against the moat-erosion bear case, Loeb makes the long-side valuation argument: NVDA is "a catch up trade... at 15 times 27, 12 times 28 for the most dominant, very fast growing company at its size." He reviewed Third Point's whole semicap/hyperscaler book expecting to take profits and instead concluded "it's the most attractive sector right now — it's where the bulk of our capital is invested," conditional only on the AI cycle not "rolling over in 31 or 32." A useful counterweight to the Feldman CUDA-erosion framing: the moat-at-inference question and the 12-month valuation question are separable.

Key facts

  • Moved from 400mm² to 800mm² die over ~5–6 years "for this exact reason" — bigger chips process more information in less time (Feldman's framing for why wafer-scale is the logical extreme). From andrew-feldman in 2026-05-21-odd-lots-why-cerebras-ceo-andrew-feldman-built-the-world-s.
  • GPU memory is HBM-class (high capacity, slow) → slow inference; the cost/power-per-token rises as you push for speed ("like miles-per-gallon falling as you drive faster"). From andrew-feldman.

  • 50% of TSMC CoWoS capacity consumed by Nvidia (~800–850K wafers/year, 2026). See cowos-packaging-capacity-crunch.

  • CEO Jensen Huang named by Feldman alongside Hock Tan (Broadcom) and Lisa Su (AMD) as great CEOs of the era.
  • Q1 FY2027 (Feb–Apr 2026, reported May 20, 2026): Revenue $82B (+85% YoY, +20% QoQ); Data Center $75B (+92% YoY, +21% QoQ) — new segmentation: Hyperscale $38B / ACIE (AI-cloud, internet, enterprise) $37B. GAAP gross margin 74.9%. Free cash flow $49B. Q2 FY2027 guide: $91B ±2%; gross margin 74.9% GAAP. Dividend raised from $0.01 → $0.25/share (25× increase). Share repurchase authorization: $80B. Sovereign AI revenue +80% YoY across ~40 countries. From 2026-05-20-earnings-nvda-q1-fy2027. (Prior autoresearch synthesis [$81.6B figure] from 2026-05-26-autoresearch-semis-ai-infra-macro-scan-may-23-26-2026 superseded by primary source.)
  • VeraRubin production shipments Q3 2026 (confirmed): jensen-huang in 2026-05-20-earnings-nvda-q1-fy2027: "VeraRubin is going to be even more successful than Grace Blackwell at this point. Every single... frontier model company will jump on VeraRubin." Standalone Vera CPU: $20B disclosed revenue visibility; the Vera CPU is a new $200B TAM opportunity.
  • LPX/SRAM inference "niche" (primary-source counter to CUDA-moat-erosion thesis): jensen-huang in 2026-05-20-earnings-nvda-q1-fy2027: "I expect that LPX and other SRAM based decode focus... accelerators. Will always be will be a niche product for some time." Basis for lowering cuda-moat-erosion-to-nvda-rerate conviction from medium → low-medium.
  • N1X ARM PC chip (Computex, June 1, 2026): Jensen Huang unveiled ARM-based N1X processor, co-developed with Microsoft and Dell, targeting Windows PCs with 14mm form factor devices, fall 2026 launch. From 2026-06-02-autoresearch-regulatory-antitrust-exec-capex-june2. Spec color from 2026-06-06-podcast-moonshots-anthropic-files-965b-ipo-trump-signs-ai-executive (panel, source-attributed): "20 CPU cores, 6,144 CUDA cores … on par with the RTX 5070 in a laptop chip"; read as a "direct shot across the bow to Apple, Intel and AMD" and an edge-intelligence / "Jarvis at home" play (paired with open-weight Cosmos/Nemotron) to block AMD's Strix Halo. The same panel asserts Nvidia is now "the majority of transistors coming out of TSMC at their bleeding edge node … no longer Apple" — unverified panel claim, but if true a supply-share datapoint for the TSMC-capacity thesis (Nvidia, not Apple, as the leading-edge anchor colliding for wafers). Flag to corroborate, not a cited fact.
  • H20 charge: $4.5B charge for H20 inventory/purchase obligations (prior period event); guided zero China DC revenue going forward. From 2026-05-22-autoresearch-china-blackwell-compute-routing-may-2026.
  • China export-control status (May 2026): US cleared ~10 Chinese firms (Alibaba, Tencent, ByteDance) for H200 purchases (BIS case-by-case review, Jan 15, 2026 policy change). But not a single H200 has shipped — Beijing blocked domestic firms from buying, steering them to Huawei Ascend. China's self-block is currently the binding constraint, not US export policy. From 2026-05-22-autoresearch-china-blackwell-compute-routing-may-2026.
  • B30A (China Blackwell): Downgraded single-die Blackwell chip for China market; $6.5–8K price target; June 2026 production target. Commercial viability unclear given China's domestic-chip-first policy. From 2026-05-22-autoresearch-china-blackwell-compute-routing-may-2026.
  • Palantir partnership / Nemotron "gloves off" (July 2026): Palantir will use Nvidia's open-source Nemotron models to build a "custom frontier quality model" for the US government ("Sovereign AI Operating System"; agencies own hardware, data, and weights). From 2026-07-03-podcast-all-in-podcast-ai-sovereignty-wars-palantir-nvidia-deal-scotus. jason-calacanis's read on the timing (uncorroborated but mechanistically coherent): Nvidia downplayed Nemotron while its top lab customers were buying GPUs, and only started promoting it "after OpenAI announced their jalapeno chips, after Anthropic started making chips, after AMD did successful projects with both of these companies, after Elon said he's going to do his own fab — Nvidia's taking the gloves off ... They are going to own the whole stack." Strategically this is the monopsony-defense david-sacks articulates in the same source: a chip company wants "as diverse and healthy an ecosystem as possible where there's lots of potential buyers for your chips," and enterprises rolling their own on open models create "a long tail of buyers." Canonical chain: frontier-lab-vertical-integration-to-sovereign-ai-stack; concept: open-source-share-shift-bullish-for-compute.

Groq deal and DOJ risk

  • Nvidia-Groq $20B deal (December 2025): Licensing + acqui-hire. Groq's LPU (Language Processing Unit) — primary inference chip architecture faster than CUDA for certain workloads — licensed to Nvidia, key personnel including CEO acqui-hired. Structured as licensing to potentially avoid HSR pre-merger review. Effect: eliminates primary inference competitor. Senators Warren/Blumenthal set April 3, 2026 deadline for Nvidia to respond; April deadline passed with no public enforcement action as of June 2, 2026. FTC separately examining AI acqui-hires. If deal ultimately blocked, NVDA loses Groq LPU access (150 TB/s on-chip SRAM, 35× energy efficiency vs Blackwell for trillion-parameter models). From 2026-05-27-autoresearch-regulatory-antitrust-semis-ai-may-2026 and 2026-06-02-autoresearch-regulatory-antitrust-exec-capex-june2.
  • DOJ Nvidia investigation (ongoing): Subpoenas issued 2024/early 2025 for market-dominance concerns (difficulty switching GPU suppliers, customer penalties for non-exclusive use, RunAI acquisition). Groq deal provides second DOJ vector. Trump DOJ posture toward tech monopoly enforcement is uncertain but has been less aggressive than Biden. Tail risk: forced Groq license divestiture, behavioral remedies, or stock uncertainty during investigation.
  • China-Russia SAMR/FAS bilateral antitrust MOU (May 25, 2026): China's State Administration for Market Regulation (SAMR) and Russia's Federal Antimonopoly Service (FAS) signed a 2026–2027 bilateral antitrust cooperation memorandum, formalizing coordinated enforcement capacity against US tech companies. Named targets: Nvidia (SAMR probe ongoing since December 2024, Mellanox acquisition condition violations, preliminary breach September 2025 — NOT closed despite Trump-Xi May 2026 summit) and Qualcomm (SAMR Autotalks acquisition investigation opened October 2025, NOT closed). From 2026-05-30-autoresearch-regulatory-antitrust-tech-biotech-utilities.

Strengths (from a thesis-input perspective)

  • Still the default for training; CUDA's training role is "shrinking" but not gone (GPT still trained in CUDA).
  • Extremely cheap per token at low speed — owns the cost-optimized end of inference.
  • Scale advantage in supply chain (CoWoS, HBM allocation).

Weaknesses (from a thesis-input perspective)

  • CUDA moat is being routed around at inference and increasingly at training (TPU, Trainium, wafer-scale).
  • Architecturally expensive/power-hungry at the fast-token end where engaged/agentic workloads are migrating.
  • Export-control exposure to China; competitor CEOs publicly disagree with Nvidia's access-maximizing stance.

Open questions

Sources

Related

Referenced by
Mechanisms
AI capex order books → non-tech industrials trade like semis (SMH-correlation screen)AI's real-world-data gap → daily-global Earth archive → Planet data moatFinancing (not chips, not datacenters) becomes the binding AI constraint → Nvidia lends its AA credit rating as a take-or-pay backstop → neoclouds become financeable → Nvidia captures an ~18% revenue share and reshapes who buys GPUsFrontier-lab vertical integration → enterprise trust break → open-model sovereign stack (Nvidia + Palantir rerate, model-layer margin pressure)Inference-demand explosion → wafer-scale fast-memory architecture → routes around HBM/CoWoS/3nm → constraint shifts to data centersSpaceX IPO passive-demand shortfall → forced Mag7 selling → equal-weight S&P outperforms
Concepts
CUDA moat erosion at inferenceInference speed as a pricing premiumOpen vs closed source model economicsOpen-source share shift is bullish, not bearish, for computeToken-price inflation favors asset-heavy compute owners
Questions
Hypothesis: AI power delivery ICs — M&A consolidation premium in MPWR/VICR/TXNHypothesis: Arm Holdings as Vera CPU architecture licensor → datacenter royalty upside not priced inHypothesis: Corning (GLW) as AI optical connectivity picks-and-shovelsHypothesis: IREN as Nvidia's Preferred DSX AI Factory Operator → Public Equity BeneficiaryIs the China-via-Malaysia Blackwell compute routing loophole durable, and does it structurally expand Nvidia's TAM?
Entities
Alex KarpAndrew FeldmanARM HoldingsBroadcomCerebrasCorning Inc.David FriedbergEtchedIREN Limited (IREN)Jan Van EckJason CalacanisMeta PlatformsSMIC (Semiconductor Manufacturing International Corporation)
Sources
Autoresearch: China Export Controls, Rare Earth Minerals & Taiwan Geopolitics (June 2026)