Market Share Distribution
Competition
March 24, 2026
Documents market share distribution across hyperscalers and neoclouds in the AI GPU cloud infrastructure segment
~63%
Big Three Combined
(AWS + Azure + GCP)
(AWS + Azure + GCP)
$23B
Neocloud Revenue
2025 Projected
2025 Projected
69%
Neocloud CAGR
→ $180B by 2030
→ $180B by 2030
+205%
Neocloud YoY Growth
Q2 2025
Q2 2025
Total Cloud Infrastructure Market Share
Q3 '25
revenue-based
AWS
29%
Microsoft Azure
20%
Google Cloud
13%
CoreWeave
~3%
Oracle Cloud
~3%
Others
32%
Neocloud Segment Revenue Distribution · 2025 Projected ($23B)
Historical Market Leadership Evolution
2006 – present
AWS
Peak ~33% (2019) → 29% (Q3 2025) · Gradual erosion
2014 – present
Microsoft Azure
13% (2018) → 20% (2025) · OpenAI partnership accelerated capture
2012 – present
Google Cloud
6% (2018) → 13% (2025) · TPU + Gemini ecosystem
2022 – present
CoreWeave
Near-zero → ~3% · Fastest cloud revenue growth in history
2022 –
Neocloud Category Emerges
GenAI demand surge created distinct GPU-first competitive tier
Recent Market Share Changes · 2024 – 2025
| Competitor | Direction | Magnitude | Key Signal |
|---|---|---|---|
| AWS | ↓ Loss | −2 pp | 31% → 29%, steady erosion |
| Azure | ↑ Gain | +1–2 pp | OpenAI workload capture |
| GCP | ↑ Gain | +1 pp | Gemini + analytics differentiation |
| CoreWeave | ↑ Gain | Significant | $465M → $1.9B → $8B (2023–25) |
| Nebius | ↑ Gain | New entrant | ARR guidance $900M–$1.1B |
| Neoclouds (agg.) | ↑ Gain | +205% YoY | >$5B Q2 2025, $23B full-year |
Competitive Dynamics · Three-Tier Market Structure
Tier 1
Hyperscalers
~63% combined share
Consolidating. Ecosystem lock-in, custom silicon (Trainium, Maia, TPU), strategic AI partnerships. Entry barrier: $50B+ annual capex.
AWS · Azure · GCP
Tier 2
Neoclouds
$23B · 30+ entrants
Fragmenting, then consolidating to 5–7 players by 2028–2030. Price advantage 50–70% vs hyperscalers. GPU rate compression: $8/hr → <$2/hr.
CoreWeave · Lambda · Nebius · Together AI · Hut8 · TeraWulf
Tier 3
Energy-Native / Sovereign
Emerging
Structural energy cost advantage + data sovereignty. No BTM hydro JV model exists. 1for.ai targets this uncrowded niche.
Crusoe · IREN · Core42 (UAE) · HUMAIN (KSA) · 1for.ai
Key Characteristics of Market Leaders
AWS
Hyperscaler
Share: 29%
Revenue: ~$115B
Regions: 30+
200+ services. H100 P5 instances + Trainium/Inferentia custom silicon. Enterprise compliance (FedRAMP, HIPAA, SOC).
Azure
Hyperscaler
Share: 20%
Regions: 60+
AZs: 116
Exclusive OpenAI partnership. Maia AI accelerator + Cobalt ARM. Hybrid cloud strength (Azure Arc/Stack).
GCP
Hyperscaler
Share: 13%
Revenue: ~$75B+
TPU v5e/v5p differentiation. Vertex AI + Gemini ecosystem. Strongest in data analytics and ML-native workloads.
CoreWeave
Neocloud
Revenue: ~$8B proj.
Backlog: $55.6B
Power: 590 MW
Kubernetes-native GPU infra. Platinum ClusterMAX. GPU-collateralized debt ($18.8B). Key clients: Microsoft, OpenAI, Meta.
Crusoe
Energy-Native
Footprint: 9.8M sq ft
Power: 3.4 GW
Vertically integrated power-to-compute. Ex-flare-gas, transitioning to broader renewable portfolio. Purpose-built for AI.
Nebius
Neocloud
ARR: $0.9–1.1B
Contracted: 2.5 GW
Ex-Yandex cloud division. Gold ClusterMAX. European data sovereignty. Connected: ~220 MW.
1for.ai Structural Position
No BTM hydro JV operator exists globally.
Tier 3 is structurally uncrowded — first mover advantage is real.
Tier 3 is structurally uncrowded — first mover advantage is real.
Energy-native · Zero-carbon · Data sovereign · Behind-the-meter hydro
