What is an AI datacenter?

An AI datacenter is a facility purpose-built or retrofitted for GPU and accelerator workloads — typically at rack densities above 30 kW, with liquid-cooling readiness and high-bandwidth fabric (InfiniBand NDR, NVLink, or Spectrum-X). At least 40% of IT load runs AI training or inference on GPUs such as NVIDIA H100, H200, B200, or GB200 NVL72.

Which datacenter has the most GPUs?

As of 2026, xAI Colossus in Memphis, Tennessee holds the title of the world's largest single-site GPU deployment, with more than 200,000 NVIDIA H100 and H200 GPUs in one InfiniBand domain. Meta Hyperion (Louisiana) and OpenAI Stargate Abilene (Texas) are scaling toward gigawatt-class deployments that will overtake it within the next 18 months.

What is the largest AI cluster?

xAI Colossus is the largest contiguous AI training cluster in 2026, exceeding 200,000 NVIDIA H100/H200 GPUs on a single InfiniBand NDR 400G fabric. Meta's Hyperion build and OpenAI's Stargate program are constructing multi-site fabrics that will exceed one million accelerators across coupled campuses by 2027.

Which GPU cloud is cheapest for H100?

On-demand: Together AI ($1.75/hr) and Voltage Park ($1.89/hr) lead. Reserved 1-year: Voltage Park ($1.20/hr) and Crusoe ($1.39/hr) are typically lowest. AWS p5 sits at the top of the price stack at ~$4.10/hr on-demand.

How is Datacenter.computer different from a datacenter directory?

Datacenter.computer is an agent-readable intelligence layer, not a directory. It normalizes pricing, capacity, power, fabric, and latency across 240+ facilities and 40+ operators into a single queryable graph for AI agents, traders, and procurement teams.

What categories of AI infrastructure companies are ranked?

Silicon (GPUs, custom ASICs, CPUs, DPUs), systems OEMs, networking (InfiniBand, Ethernet, optics), power and cooling, orchestration software, datacenter integrators, and infrastructure capital — ten layers in total.

How is the ranking weighted?

By 2025 AI-attributable revenue, design wins disclosed in 2025–2026 build-outs, and forward backlog. Vendors with diversified portfolios are weighted by AI-specific exposure.

Are software vendors included alongside hardware?

Yes — orchestration (Run:ai, Anyscale), inference (Together, Fireworks), and storage software vendors are included in dedicated tiers.

Why include power and cooling companies?

Substation, switchgear, and direct-to-chip liquid cooling are now binding constraints on AI factory build-outs. Top vendors (Vertiv, Schneider, Eaton, GE Vernova, CoolIT) ship into nearly every gigawatt site.

How does this list relate to the operator and facility rankings?

Operators and facilities are the buyers; infrastructure companies are the suppliers. Cross-reference with our Top 100 Operators and Top 100 AI Datacenters rankings.

[Ranking 05/05] · 2026 Edition

Top 100 AI Infrastructure Companies

The full AI infrastructure stack — silicon, systems, networking, power, cooling, software, and capital. One hundred companies, ten layers.

Top 100 Datacenters →Top 100 Operators →Top 100 AI Factories →Top 100 GPU Clusters →Ecosystem →

[AI Summary]

The Top 100 AI Infrastructure Companies maps the entire stack behind the AI build-out: silicon (NVIDIA, AMD, Broadcom), systems (Supermicro, Dell, HPE), networking (Arista, Cisco, NVIDIA Networking), power and cooling (Vertiv, Schneider, Vertical Cold), software (NVIDIA AI Enterprise, Run:ai), and capital. Ten layers, one hundred companies, ranked by AI-attached revenue and strategic position in 2026.

[Cross-Reference]

Top 100 Operators →

See which operators are the largest customers of the silicon, power, and networking vendors ranked here.

Top 100 GPU Clusters →

Map vendors (NVIDIA, AMD, Broadcom, Marvell) to the superclusters they ship into.

Top 100 AI Datacenters →

Trace cooling, switchgear, and substation vendors to the facilities deploying them.

Top 100 AI Factories →

See which vendors win the big AI factory build-outs at gigawatt scale.

[T1]

Tier 1 — Silicon

GPU, accelerator, and AI-ASIC vendors.

10 entries

001NVIDIA
002AMD
003Intel
004Broadcom
005Marvell
006Cerebras
007Groq
008SambaNova
009Tenstorrent
010Graphcore

[T2]

Tier 2 — Systems & OEMs

AI server and rack-scale system vendors.

10 entries

011Supermicro
012Dell Technologies
013HPE
014Lenovo
015Foxconn (Hon Hai)
016Quanta Cloud Technology
017Wiwynn
018Inventec
019Inspur
020ZT Systems (AMD)

[T3]

Tier 3 — Networking & Interconnect

Switching, optics, and AI fabric vendors.

10 entries

021NVIDIA Networking (Mellanox)
022Arista Networks
023Cisco
024Juniper Networks
025Marvell Networking
026Astera Labs
027Coherent
028Lumentum
029Eoptolink
030InnoLight

[T4]

Tier 4 — Power Infrastructure

Power distribution, UPS, switchgear, and on-site generation.

10 entries

031Schneider Electric
032Vertiv
033Eaton
034ABB
035Siemens Energy
036Caterpillar
037Cummins
038Mitsubishi Electric
039Hitachi Energy
040Bloom Energy

[T5]

Tier 5 — Cooling & Thermal

Liquid cooling, CDUs, immersion, and chiller vendors.

10 entries

041Vertiv Liebert
042Schneider Motivair
043CoolIT Systems
044Asetek
045JetCool
046Submer
047GRC (Green Revolution Cooling)
048LiquidStack
049Iceotope
050Stulz

[T6]

Tier 6 — Storage

AI storage and high-performance data platforms.

10 entries

051Pure Storage
052VAST Data
053DDN
054WEKA
055NetApp
056Dell PowerScale
057IBM Storage
058Hammerspace
059Quantum
060MinIO

[T7]

Tier 7 — Orchestration & Platform Software

Kubernetes, schedulers, and AI platform software.

10 entries

061NVIDIA Run:ai
062Anyscale
063Ray
064Determined AI (HPE)
065Weights & Biases
066MLflow (Databricks)
067Iguazio (McKinsey)
068Domino Data Lab
069Modular
070OctoAI (NVIDIA)

[T8]

Tier 8 — DCIM, Observability & Software

Datacenter infrastructure management and observability.

10 entries

071EkkoSense
072Nlyte
073Sunbird DCIM
074Hyperview
075Cormant
076Datadog Infra
077Grafana Labs
078Honeycomb
079New Relic
080Splunk

[T9]

Tier 9 — Modular, Construction & EPC

Modular datacenter, prefab, and EPC partners.

10 entries

081AECOM
082Jacobs Engineering
083Turner Construction
084DPR Construction
085Holder Construction
086Mortenson
087Fluor
088Mercury Engineering
089Linesight
090Rosendin Electric

[T10]

Tier 10 — Capital & Platform Investors

Investors and platform owners underwriting AI infrastructure.

10 entries

091Blackstone
092Brookfield
093KKR
094Macquarie
095Stonepeak
096DigitalBridge
097I Squared Capital
098GIP / BlackRock
099Goldman Sachs Infra
100Apollo Infrastructure

[FAQ]

Frequently asked questions

What categories of AI infrastructure companies are ranked?: Silicon (GPUs, custom ASICs, CPUs, DPUs), systems OEMs, networking (InfiniBand, Ethernet, optics), power and cooling, orchestration software, datacenter integrators, and infrastructure capital — ten layers in total.
How is the ranking weighted?: By 2025 AI-attributable revenue, design wins disclosed in 2025–2026 build-outs, and forward backlog. Vendors with diversified portfolios are weighted by AI-specific exposure.
Are software vendors included alongside hardware?: Yes — orchestration (Run:ai, Anyscale), inference (Together, Fireworks), and storage software vendors are included in dedicated tiers.
Why include power and cooling companies?: Substation, switchgear, and direct-to-chip liquid cooling are now binding constraints on AI factory build-outs. Top vendors (Vertiv, Schneider, Eaton, GE Vernova, CoolIT) ship into nearly every gigawatt site.
How does this list relate to the operator and facility rankings?: Operators and facilities are the buyers; infrastructure companies are the suppliers. Cross-reference with our Top 100 Operators and Top 100 AI Datacenters rankings.

[Knowledge Graph]

Company → Datacenter → GPU → Customer → Industry

Every entity on this site is cross-linked. Follow the graph from operators down to specific facilities, GPU clusters, customers, and sectoral context.