ThematIQ

AI Data Center Build-out

Everything it takes to build and run an AI data center.

The market$1.7 trillion

Worldwide data center capex, all data center IT and infrastructure, not only AI

Dell'Oro Group, 2030
The build18

parts in the AI Supercycle bill of materials, each traced to the companies that supply it

ThematIQ research
The companies86

companies worth $17.9 trillion, earning on average 66% of their revenue here

Annual reports and exchange data, 2026-10-28
Build a strategy from AI Data Center Build-out

What it holds

AI Data Center Build-out follows the money through the AI infrastructure bill of materials: the parts it takes to build it, and the companies that earn their revenue supplying them. Each company is read from its own reports, so a holding reflects what it actually sells into the theme, not what it says about it.

PartWhat it covers
Graphics Processing Unit (GPU)GPU accelerators, AI ASICs custom, foundry leading edge, advanced packaging
Video Memory (VRAM)HBM memory, GDDR memory, DRAM
Data Processing Unit (DPU)DPU SmartNIC
Cooling Systemdirect to chip cooling, thermal management components
Voltage Regulator Module (VRM)voltage regulator modules, power management ICs
Motherboard Interfaceserver motherboards, PCIe CXL retimers switches
Interconnection Interfaceinfiniband, ethernet switch silicon, optical transceivers, SerDes ip, interconnect ICs
Network Interface and ControllerNIC controllers
Output Interfacesdisplay interface ICs
Software (AI hardware performance)AI hardware software stack, EDA software
Power and Cooling Systemsdata center power distribution, UPS systems, power supply units, liquid cooling, data center thermal systems
Servers and IT Equipmentservers AI, server CPUs, server ODM
Networking Equipmentdata center switches routers, optical networking equipment
Storage Devicesenterprise SSD, NAND flash, storage controllers, HDD nearline
Backup and Disaster Recovery Solutionsbackup recovery software
Racks and Cable Managementserver racks enclosures, data center cabling connectivity
Remote Hands Supportdata center onsite services
Datacenter Softwarevirtualization software, data center infrastructure management, data center security software

How it breaks down

By listing country
United States: 47%Taiwan: 32%China: 13%France: 4%South Korea: 3%
  • United States47%
  • Taiwan32%
  • China13%
  • France4%
  • South Korea3%
By company size (market value)
Mega cap (over $200bn): 32%Large cap ($10bn to $200bn): 45%Mid cap ($2bn to $10bn): 22%Small cap (under $2bn): 1%
  • Mega cap (over $200bn)32%
  • Large cap ($10bn to $200bn)45%
  • Mid cap ($2bn to $10bn)22%
  • Small cap (under $2bn)1%

Shares of the strategy as of the data date.

Largest companies

CompanyListingMarket value
Auras Technology3324.TWO$4 billion
Modine ManufacturingMOD.XNYS$8 billion
Advanced Energy IndustriesAEIS.XNAS$10 billion
NvidiaNVDA.XNAS$5.2 trillion
Taiwan Semiconductor Manufacturing2330.TW$1.9 trillion
ASE Technology Holding3711.TW$95 billion
Vertiv HoldingsVRT.XNYS$94 billion
Marvell TechnologyMRVL.XNAS$221 billion
LegrandLR.PA$39 billion
Delta Electronics2308.TW$124 billion
Asia Vital Components3017.TW$39 billion
Advanced Micro DevicesAMD.XNAS$944 billion
BroadcomAVGO.XNAS$1.7 trillion
Shenzhen Envicool Technology002837.SHE$6 billion
Micron TechnologyMU.XNAS$1.1 trillion

The market

$1.7 trillion

Worldwide data center capex, all data center IT and infrastructure, not only AI

Dell'Oro Group, 2030
$580bn

Global investment in data centers as the IEA defines it, not AI-only

IEA, as reported by Data Center Knowledge, 2025

The Dell'Oro forecasts were revised upward between Feb 2026 and Aug 2026, and the IEA figure is a 2025 investment number from an energy agency, so it is not directly comparable to the 2030 forecasts. The note gives no growth rates for the whole segment, and its component-level figures are generic-vendor estimates that disagree by large factors, so I left them out.

All 4 published estimates, with sources
What it measuresFigureYearSource
Worldwide data center capex, all data center IT and infrastructure, not only AI$1.7 trillion2030Dell'Oro Group
Worldwide data center capex, all data center IT and infrastructure, not only AI (expected to approach this level)$1.0 trillion2026Dell'Oro Group
Worldwide data center capex, same scope, revised upward (surpasses this level)$3.0 trillion2030Dell'Oro Group
Global investment in data centers as the IEA defines it, not AI-only$580bn2025IEA, as reported by Data Center Knowledge

Questions

What is AI Data Center Build-out?
Everything it takes to build and run an AI data center.
What does AI Data Center Build-out hold?
Listed companies that earn their revenue from the parts it takes to build AI Supercycle: Graphics Processing Unit (GPU), Video Memory (VRAM), Data Processing Unit (DPU), Cooling System, Voltage Regulator Module (VRM), Motherboard Interface. 86 companies as of 2026-10-28.
What is not in AI Data Center Build-out?
Cold plates, vapor chambers, thermal interface materials and server level liquid loops: Cooling System, not Power and Cooling Systems; CDUs, rack manifolds, rear door exchangers, chillers and CRAH units: Power and Cooling Systems, not Cooling System Standard and RDMA NIC adapters and controllers: Network Interface and Controller, not Networking Equipment; programmable SmartNICs and DPUs with onboard compute cores: Data Processing Unit (DPU), not Network Interface and Controller Pluggable optical transceivers, AOCs, DACs and AECs: Interconnection Interface, not Networking Equipment or Racks and Cable Management; long haul and DCI optical transport systems: Networking Equipment; passive fiber trunks, patch panels and structured cabling: Racks and Cable Management

Sources

Every part of the bill of materials is researched from published industry sources, and every company from its own annual reports and exchange data. Nothing is estimated that cannot be sourced.

Industry research: acecampai.substack.com/p/superpods-are-not-bigger-clusters · ai-stack.us/products/2-248-nvidia-connectx-8-supernic-900-9x · akashborate.substack.com/p/understanding-how-to-bring-up-nvi · arxiv.org/pdf/2311.02651 · blog.barrack.ai/nvidia-b300-1400w-data-center-requirements/ · builtin.com/articles/optimize-data-backup-recovery-costs · capitalandcompute.net/memory-prices/hbm/ · convergedigest.com/delloro-ai-back-end-switch-market-to-near · dev.to/gentic_news/nvidia-feynman-gpu-power-semi-content-hit · docs.nvidia.com/enterprise-reference-architectures/nvl72-ai- · electronics.alibaba.com/buyingguides/42u-server-rack-price-g · epoch.ai/data-insights/b200-cost-breakdown · finance.biggo.com/news/X4vnUp4B2jrwCtglf11J · finance.yahoo.com/news/cooling-system-single-nvidia-blackwel · finance.yahoo.com/real-estate/articles/data-center-construct · finance.yahoo.com/technology/ai/articles/nvidia-already-make · fourweekmba.com/ai-nvidias-huang-pegs-software-at-about-2000 · fourweekmba.com/hbm-the-memory-that-controls-45-of-gpu-econo · fourweekmba.com/nvidia-compute-revenue/ · github.com/bmsdebug/chinascrape/issues/23

What it is not

Where an item could belong to two parts, these rules decide which owns it; items outside every part are not counted.

Not investment advice. Figures are dated and sourced; ThematIQ does not estimate figures it cannot source.