Token Runner
SPRAWL // 03 / INDUSTRY

INDUSTRY

Lab consolidation, supply chain bottlenecks, hardware allocations

DATA • CONVERSATIONS • SIGNALS • HUMANS • MACHINES

65%of AI compute sold was Nvidia’s (Epoch AI’s estimate)2025
Industry: a painting from the house’s oil series
01 /Interlinked6pieces, each an article and a film02 /Data3charts, each live from a table we hold03 /Conversations8posts this period and quoted04 /Signals22sources our pieces cite05 /Humans66people on the Replicants list06 /Machines28chip types sold, by Epoch AI’s estimate

Interlinked

Industry
Industry

Data

OPEN THE INSTRUMENT →

Open Industry in Esper →

The rows behind this figure: AI chips sold in 2025, by chip and quarter (Epoch AI’s estimates)

The table is read from the same data file as the chart when this section is opened.

All Interlinked: Industry

Industry
Industry
Politics interlinked
Politics interlinked
Economics interlinked
Economics interlinked

Conversations

The biggest conversational points among Industry’s tracked accounts and searches since 2026-10-17: 87 posts read, 22 kept, 45 dropped as not about AI, 3 dropped as a promotion, 12 dropped as only a link, 5 dropped as a reply to a reply. The posts are their authors’ own words; the rule under each pick says why it stands here.

1. NVIDIA

2 accounts · 3 posts · engagement 5,182 · the point they share: a lane term several carry

SemiAnalysis@SemiAnalysis_OPEN THE POST ON X →

BREAKING NEWS🚨: NVIDIA has just achieved 423 tok/s/user 🚀 on the 2.8T Parameter Kimi K3 model via a new upstream vLLM optimization. 1.68x Faster on the Same Hardware!

This was done by NVIDIA CentML's Benjamin, who wrote NVIDIA's new decode-specialized kernels in Cute DSL via aggressive Megakernel fusion. This is a 12,000 LoC Megakernel and is the future of advanced AI-assisted kernel generation. NVIDIA is already working on DeepSeekv4.1 and other models for vLLM too.

Someone should give this NVIDIA 10x engineer a raise for this great new major feature!

AS POSTED ON X
Why this post

the most engaged post of @SemiAnalysis_ in this conversation (2063 likes, 154 reposts, 61 quotes, 118 replies as read); the point: a lane term several carry (NVIDIA); from the lane

NVIDIA@nvidiaOPEN THE POST ON X →

AI builders need computing capacity they can put to work sooner.

We introduced DSX POD at NVIDIA GTC Berlin: AI factories integrated and tested before delivery, enabling deployment at prepared sites in as little as 14 days. pic.x.com/0Ax7yHMvBp

AS POSTED ON X
Why this post

the most engaged post of @nvidia in this conversation (1507 likes, 167 reposts, 78 quotes, 131 replies as read); the point: a lane term several carry (NVIDIA); from the lane

2. infrastructure

2 accounts · 2 posts · engagement 459 · the point they share: a lane term several carry

AMD@AMDOPEN THE POST ON X →

Agentic AI demands a new kind of infrastructure.

At #OCPSummit26 we announced AMD Helion, bringing an open rack-scale design for CPU-centric agentic AI deployments, with flexible configurations for enterprises.

Full story on @TomsHardware: bit.ly/4ypGxMq

AS POSTED ON X
Why this post

the most engaged post of @AMD in this conversation (225 likes, 24 reposts, 3 quotes, 25 replies as read); the point: a lane term several carry (infrastructure); from the lane

Intel@intelOPEN THE POST ON X →

Every prompt has its place. So how do we match each AI request to the right model and compute?

Watch Nicolas Dube’s #OCPSummit26 keynote, “Put Every Prompt in Its Place,” to explore the role of open infrastructure in scaling AI: ms.spr.ly/6012avVQE pic.x.com/mpMNVjbzGP

AS POSTED ON X
Why this post

the most engaged post of @intel in this conversation (91 likes, 8 reposts, 3 quotes, 10 replies as read); the point: a lane term several carry (infrastructure); from the lane

The posts our Industry pieces quoted

Not a feed. Each with the piece that used it.

Cassandra Unchained@michaeljburryOPEN THE POST ON X →

When the circularity questions got loud November of 2025, Nvidia’s formal rebuttal was that its “strategic” investments in its customers ‘represent a minimal portion of revenue’ already booked. Granted. However, BIS’s chart is about the announced multi-year commitments. This is the forward book, part of Jensen’s “visibility” to $1 trillion. There, circular financing is nearly 100% to date.
Short Thoughts - NVDA, Neos, Hyperscalers, Jevons Paradox, and Compression open.substack.com/pub/michaeljbu…

AS POSTED ON X

Quoted in Who holds the Nvidia chips

Cassandra Unchained@michaeljburryOPEN THE POST ON X →

That $500 billion $NVDA Wall Street stunt involves Nvidia taking 25% stakes & providing residual value guarantees on purchase of its chips. All filtered through Private Equity’s Private Credit schemes. I have an idea how that will look. Meet the new Boss. Same as the old Boss https://t.co/vzAnWLkhxI

AS POSTED ON X

Quoted in Who holds the Nvidia chips

Epoch AI@EpochAIResearchOPEN THE POST ON X →

For a sense of scale, this trade could represent 150,000 H100e of compute, about a quarter of our earlier modeled estimate of all chips smuggled into China through 2025.

Read our previous full report on our modeled estimates: epoch.ai/publications/c…

AS POSTED ON X

Quoted in Who holds the Nvidia chips

CSET@CSETGeorgetownOPEN THE POST ON X →

#AI chip smuggling is like an iceberg: we only see what’s above the surface. To estimate its true scale, we need to know how much is hidden below the waterline. Whether that’s 50% or 90% makes a big difference. In this Q&A, @jacob_feldgoise describes how location verification methods can—and cannot—improve enforcement of export controls on these powerful chips. Link below ⬇️

AS POSTED ON X

Quoted in Who holds the Nvidia chips

Signals

The filings, reports and news stories behind the Industry pieces, as each piece cites them. A document two pieces cite shows twice.

Humans

The people of the 2026 Replicants list who sit in Industry: 66, by rank. A card opens the person on the Replicants page.

#3Jensen HuangFounder, president and chief executive, NVIDIAChips#5Satya NadellaChairman and chief executive, MicrosoftHyperscalers#9Sundar PichaiChief executive of Google and AlphabetHyperscalers#15Tim CookExecutive chairman, AppleHyperscalers#17Larry EllisonExecutive chairman and chief technology officer, OracleHyperscalers#20Jeff BezosFounder of Amazon; co-chief executive of PrometheusEconomics interlinkedCapital#24Lisa SuChair and chief executive, AMDChips#27Andy JassyPresident and chief executive, AmazonHyperscalers#32C.C. Wei (Wei Zhejia)Chairman and chief executive, TSMCChips#37Dylan PatelFounder, chief executive and chief analyst, SemiAnalysisContent producers#38Hock TanPresident and chief executive, BroadcomChips#43Ben ThompsonFounder and author, StratecheryContent producers#45Kevin ScottChief technology officer, MicrosoftHyperscalers#62Thomas KurianChief executive, Google CloudHyperscalers#63Matt GarmanChief executive, Amazon Web ServicesHyperscalers#64Michael IntratorCo-founder, chairman and chief executive, CoreWeaveChips#65Ali GhodsiCo-founder and chief executive, DatabricksHyperscalers#66Marc BenioffCo-founder, chair and chief executive, SalesforceHyperscalers#67Chuck RobbinsChair and chief executive, CiscoHyperscalers#68Jayshree UllalChair and chief executive, Arista NetworksHyperscalers#69Michael DellFounder, chairman and chief executive, Dell TechnologiesHyperscalers#93Palmer LuckeyFounder, Anduril IndustriesEconomics interlinkedCapital#94Trae StephensPartner, Founders Fund; co-founder and executive chairman, AndurilEconomics interlinkedCapital#96Rene HaasChief executive, ArmChips#97Christophe FouquetPresident and chief executive, ASMLChips#98Colette KressExecutive vice president and chief financial officer, NVIDIAEconomics interlinkedChips#99Sanjay MehrotraChairman and chief executive, Micron TechnologyChips#100Amy HoodExecutive vice president and chief financial officer, MicrosoftEconomics interlinkedHyperscalers#101Ruth PoratPresident and chief investment officer, Alphabet and GoogleEconomics interlinkedHyperscalers#102Safra CatzExecutive vice chair of the board, OracleHyperscalers#103Brad SmithVice chair and president, MicrosoftPolitics interlinkedPolicy / Politics#114Andrew BosworthChief technology officer and head of Reality Labs, MetaHyperscalers#115Chris CoxChief product officer, MetaHyperscalers#116Swami SivasubramanianVice president, Agentic AI and Emerging Technologies, Amazon Web ServicesHyperscalers#117Werner VogelsVice president and chief technology officer, AmazonHyperscalers#118John GiannandreaFormer SVP, Machine Learning and AI Strategy, Apple (retired)Hyperscalers#119Craig FederighiSenior vice president, software engineering, AppleHyperscalers#120Eddie WuChief executive officer, Alibaba GroupHyperscalers#121Joe TsaiChairman, Alibaba GroupHyperscalers#122Robin LiCo-founder and chief executive, BaiduHyperscalers#123Pony Ma (Ma Huateng)Co-founder, chairman and chief executive, TencentHyperscalers#124Zhang YimingFounder (stepped down as CEO in 2021), ByteDanceHyperscalers#125Liang RuboCo-founder, chairman and chief executive, ByteDanceHyperscalers#126Lei JunFounder, chairman and chief executive, XiaomiHyperscalers#127Ian BuckVice president, hyperscale and high-performance computing, NVIDIAChips#128Forrest NorrodExecutive vice president and general manager, Data Center Solutions Business Group, AMDChips#129Victor PengChief executive officer, PsiQuantumChips#130Mark LiuFounder, TCIP Center at UC Berkeley; former chairman of TSMCChips#131Morris ChangFounder of TSMC (retired)Chips#132Charlie KawwasPresident, Semiconductor Solutions Group, BroadcomChips#133Pat GelsingerGeneral partner, Playground Global; former CEO of IntelChips#134Lip-Bu TanChief executive officer, IntelChips#135Michelle Johnston HolthausFormer CEO of Intel ProductsChips#136Peter WenninkChairman of the board, EuclydChips#137Brian VenturoCo-founder and chief strategy officer, CoreWeaveChips#138Arkady VolozhFounder and chief executive, Nebius GroupChips#139Stephen BalabanCo-founder and chief technology officer, LambdaChips#140Chase LochmillerCo-founder and chief executive, CrusoeChips#141Cristiano AmonPresident and chief executive, QualcommChips#142David ZinsnerExecutive vice president and chief financial officer, IntelEconomics interlinkedChips#143Jeetu PatelPresident and chief product officer, CiscoChips#144Andy BechtolsheimCo-founder and chief architect, Arista NetworksChips#145Antonio NeriPresident and chief executive, Hewlett Packard EnterpriseHyperscalers#146Bill McDermottChief executive officer, ServiceNowChips#147Sridhar RamaswamyChief executive officer, SnowflakeChips#148Matei ZahariaCo-founder and chief technology officer, DatabricksChips

Open Industry on Replicants →

Machines

28chip types sold, by Epoch AI’s estimate
178machine-learning hardware releases
OPEN THE INSTRUMENT →