NVIDIA
playerAI & Compute · Money & Finance · Nations & Geopolitics
The forge — whoever controls these chips controls who's even allowed to build AI.
Who they are
NVIDIA[1].
What they do
The chokepoint of AI compute. Its GPUs carry 36-52 week wait times, so demand runs through it.
How it works
Sells a 'Sovereign AI OS' with Palantir — turnkey datacenters. Its compute is legally fenced inside the US defense perimeter through the CHIPS Act and export controls, and a co-founder co-chairs a government-industry AI task force.
Why it matters
Control the forge and you control who can build advanced AI at all — which is exactly why it sits inside the defense perimeter.
What's new (Aug–Oct 2026)
Around Aug 19–21, 2026 — Small batches of Nvidia H200 chips were allowed into mainland China (Business Standard, 19 Aug). ByteDance and Tencent each took delivery of roughly 10,000 H200s, with each purchase approved by China's planning agency, the NDRC (Tom's Hardware, 21 Aug). (business-standard.com ; tomshardware.com)
Around Aug 24, 2026 — Taiwanese prosecutors charged nine people, including one from Nvidia and two from Super Micro; 74 servers reached China. (theguardian.com)
Aug 26, 2026 — Nvidia reported revenue of $96.2bn for the quarter ended July 26, 2026, up 18% from the prior quarter and 106% from a year earlier (the Guardian said $92bn was expected). Data-center revenue was $89bn and operating income was $63.7bn (ynet). (nvidianews.nvidia.com ; ynetnews.com ; theguardian.com)
Aug 26, 2026 — Amazon and Nvidia expanded their partnership to add another 2 million Nvidia GPUs to Amazon data centers. (techcrunch.com)
Aug 28, 2026 — Reports said the Trump administration was preparing a rule aimed at Chinese access to remote AI servers, such as renting Nvidia chip time through Thailand and Singapore. It was a draft and not confirmed by the Commerce Department. (tomshardware.com ; techtimes.com)
Aug 31, 2026 — Nvidia is investing $3.5bn in MediaTek. (techcrunch.com)
Sept 3, 2026 — Nvidia agreed to buy Hugging Face for $12,930,300,000, Jensen Huang wrote on Nvidia's blog; the BBC put it at about $12.9bn (£9.5bn). Huang said people will not be required to use Nvidia computing to build on or deploy through Hugging Face. (cnbc.com ; blogs.nvidia.com ; bbc.co.uk ; techcrunch.com)
Sept 9, 2026 — The research group C4ADS published its report "Covert Compute", dated September 9, 2026. (c4ads.org)
Sept 28, 2026 — Nvidia said Anthropic has contracted $180bn in infrastructure value, and Nvidia sped up its stock buyback by $150bn to a total of $235bn. TechTimes put its Vera Rubin systems at "$40 billion per gigawatt". (techtimes.com ; gurufocus.com)
The engine's record — word for word
The Forge. Controls GPU supply chain (36-52 week lead times). INL partnership. Mar 12: "Sovereign AI OS" with Palantir — Blackwell Ultra turnkey datacenters. Genesis compute infrastructure now commercially deployable.
**R93 + R95 IC-AI substrate context (May 22 2026 backfill):** Per
Report #93: chips_act_2022_pl117_167 (PL 117-167 $52.7B + Technology Clawback) + bis_export_controls_oct_7_2022 (Federal Register 2022-21658 + Macau 2023 extension) statutorily enclose NVIDIA compute substrate inside US defense perimeter — Commerce + IC licensing oversight. Per
Report #95: SCSP-NVIDIA 'Task Force on AI and the Future of Work' March 2026 (scsp_nvidia_task_force_2026 node) — NVIDIA Co-Founder Chris Malachowsky co-chairs alongside SCSP CEO Bajraktari + Senators Mark Warner + Mike Rounds + OpenAI's Chan Park + former Commerce Secretary Gina Raimondo. Aligned Data Centers $40B AIP take-private Oct 2025 (aligned_data_centers node) provides 5GW NVIDIA H100 + Blackwell hyperscale capacity — sovereign-wealth-funded Physical-AI substrate per divergence #156 AIP/BlackRock-GIP + #169 SWF Apex-Parent Array.
**R96 Quantum 2026 9-firm context (May 22 2026 backfill):** Per
Report #96 H_3 Quantum 2026 framing: quantum_equity_2026_9_firm_allocation (R96 node, NIST + Department of Commerce primary verified) deploys $2B across IBM $1B + GlobalFoundries $375M + Rigetti + D-Wave + PsiQuantum + Quantinuum + Atom Computing + Infleqtion $100M each + Diraq $38M — operates as ALTERNATIVE-architecture compute-substrate deployment parallel to NVIDIA classical-GPU compute oligopoly. Combined classical-GPU + quantum-computing federal-capital architecture: CHIPS Act 2022 PL 117-167 (R93) + BIS Oct 7 2022 export controls (R93) + Quantum 2026 federal-equity (R96) = triple-architecture compute-substrate enclosure per
Report #96. Per SCSP-NVIDIA Task Force March 2026 (scsp_nvidia_task_force_2026 R95 node) — NVIDIA Co-Founder Chris Malachowsky co-chairs alongside Bajraktari + Senators Warner + Rounds + Chan Park (OpenAI) + Gina Raimondo, demonstrating formal federal-policy-NVIDIA coordination at task-force tier. [Report #154] (Report #154) NVIDIA's true moat is the INTERCONNECT, not compute: NVLink/NVSwitch (NVLink 5.0 = 1.8 TB/s, up to 576 GPUs as one domain) is the proprietary data-transit fabric that gates multi-chip scaling (interconnect_binding_constraint). The chokepoint the Big-Three-owned apex controls — whoever owns the interconnect owns the substrate.
[Live pass Oct 6 2026] ~2026-08-24 — Taiwanese prosecutors charged nine people (Monday) including one from Nvidia and two from Super Micro; 74 servers reached China. (1 source) (
theguardian.com) | 2026-08-26 — Nvidia Q2 FY2027 (quarter ended July 26, 2026): revenue $96.2bn, up 18% q/q and 106% y/y (Guardian: vs $92bn expected); data-center revenue $89bn; operating income $63.7bn (ynet). (
nvidianews.nvidia.com ;
ynetnews.com ;
theguardian.com) | 2026-08-26 — Amazon-Nvidia expanded partnership to add another 2 million Nvidia GPUs to Amazon data centers. (1 source) (
techcrunch.com) | 2026-08-28 — Reports: Trump administration preparing a rule targeting Chinese access to remote AI servers (renting Nvidia GPU time via Thailand and Singapore). Draft, not confirmed by Commerce. (
tomshardware.com ;
techtimes.com) | 2026-08-31 — Nvidia investing $3.5bn in MediaTek. (1 source) (
techcrunch.com) | 2026-09-03 — Nvidia agreed to acquire Hugging Face for $12,930,300,000 (Nvidia blog, 3 Sep, Jensen Huang); BBC: about $12.9bn (£9.5bn). Huang: Nvidia compute will not be required to build on or deploy through Hugging Face. (
cnbc.com ;
blogs.nvidia.com ;
bbc.co.uk ;
techcrunch.com) | 2026-09-09 — The C4ADS report page (c4ads.org/reports/covert-compute) carries the date "September 9, 2026". (1 source) (
c4ads.org) | 2026-09-28 — Nvidia: Anthropic has contracted $180bn in infrastructure value; Nvidia accelerated its buyback by $150bn to a total authorization of $235bn. TechTimes gives Vera Rubin at "$40 billion per gigawatt". (
techtimes.com ;
gurufocus.com)
[Live pass Oct 6 2026] News at odds with this record's text "statutorily enclose NVIDIA compute substrate inside US defense perimeter": ~2026-08-19 — Small batches of Nvidia H200 chips allowed into mainland China (Business Standard 19 Aug); ByteDance and Tencent each took delivery of roughly 10,000 H200s under NDRC per-purchase approval (Tom's Hardware 21 Aug). (
business-standard.com ;
tomshardware.com)
Follow the trail
Walk this on the live map →