Tesla operates the largest real-world AI inference network through FSD. Robotaxi launches in live cities shift autonomy from theory to deployment. Energy profits reached $1.1B, funding autonomy investment. Dojo and custom inference chips reduce external compute dependence as fleet-scale AI expands.
$TSLA | +28.0% YTD
Azure demand remains supply-constrained, reinforcing pricing discipline and margin durability.
$MSFT | +16.6% YTD
Microsoft anchors enterprise AI spend through Azure and OpenAI. Contracted AI cloud commitments now exceed $250B, creating long-term visibility. Copilot monetization embeds AI directly into Office workflows, raising switching costs.
Meta commoditizes the model layer with open-source Llama while monetizing outcomes through ads. AI-driven ad automation reached a $60B run rate. ARPP climbed to $14.46, reflecting efficiency gains rather than ad load. User-scale creates a feedback loop competitors cannot replicate.
$META | +11.4% YTD
Search integrations expanded query volume, supporting ad monetization.
$GOOGL | +65.8% YTD
Google controls silicon, infrastructure, models, and users. A $155B cloud backlog signals long-duration AI demand. Gemini reduced inference latency by 40%, enabling real-time multimodal use. 650M Gemini MAUs convert infrastructure advantage into distribution leverage.
Blackwell accelerated demand, turning memory from commodity into strategic asset.
$MU | +228.3% YTD
Micron sits at the AI memory choke point. HBM3e bandwidth defines system-level performance as compute scales faster than memory. Micron sold out 2025 and 2026 HBM capacity, locking in pricing leverage uncommon for cyclical memory.
Every major AI chip vendor depends on TSMC’s yields, scale, and advanced-node execution.
$TSM | +48.2% YTD
TSMC remains the unavoidable toll booth for advanced AI silicon. CoWoS packaging constrains accelerator supply, reinforcing pricing power. 3nm represents 23% of wafer revenue, and $42B in CAPEX secures future dominance.
Commercial adoption now drives growth alongside government demand.
Palantir transitioned from services-heavy deployments to repeatable software scale. AIP bootcamps compress sales cycles and convert pilots into production rapidly. 204 contracts above $1M highlight enterprise commitment. Margins expanded past 50%, driven by software leverage rather than headcount.
$PLTR | +158.2% YTD
Custom accelerators plus interconnects create a compounding revenue stream tied directly to AI cluster expansion.
$AVGO | +51.0% YTD
Broadcom benefits as hyperscalers design custom silicon to reduce Nvidia dependence. A $73B AI backlog reflects multi-year ASIC and networking demand. Ethernet-based AI back-end networks favor Broadcom’s switch portfolio as cluster size grows.
Inference economics, not training cycles, anchor durability as AI workloads scale across enterprises.
$NVDA | +36.4% YTD
Nvidia’s moat moved from chips to systems. Blackwell and GB300 push customers into rack-scale architectures, increasing switching costs beyond GPUs. Networking revenue reached $8.2B, doubling as clusters densify. CUDA remains the gatekeeper.
Paris cluster launch and a $1B European rollout targeting H200 and Blackwell GPUs create regional scarcity leverage outside US hyperscalers.
Nebius is emerging as a “sovereign AI” neocloud. Purpose-built GPU clusters, not legacy cloud baggage. A $17.4B GPU capacity agreement with Microsoft plus a $3B Meta deal locked in demand.
Top 10 Ai Companies: Where Money Is Really Going 🧵
Capital is flowing toward companies controlling bottlenecks: compute, memory, software, or distribution. YTD performance reflects where leverage sits today. 👇
$NBIS | +198.7% YTD
Rising RPO highlights long-dated commitments as customers embed Snowflake deeper into core data infrastructure.
Snowflake monetizes data gravity by bringing AI workloads directly to where data lives. Separation of compute and storage enables flexible scaling and cost control. Cortex and Snowpark allow enterprises to deploy AI models without data movement, reducing friction and risk.
$SNOW | NTM revenue growth +24.7%
The Max subscription tier lifts ARPU through GenAI features while maintaining retention, extending the flywheel beyond language learning.
$DUOL | NTM revenue growth +24.9%
Duolingo’s competitive edge comes from data scale and engagement. Billions of daily learning interactions continuously retrain AI models competitors cannot replicate. Expansion into math and music reuses the same engagement engine.
Expanding enterprise penetration and rising ARR per customer reinforce visibility and long-duration revenue characteristics.
$RBRK | NTM revenue growth +25.2%
Rubrik defines cyber resilience rather than simple backup. Regulatory pressure around breach disclosure elevates instant recovery to a board-level priority. Subscription growth is driven by ransomware defense and cloud migration, not discretionary IT upgrades.
Software and financial services drive margin expansion as international and offline GMV scale on the same core platform.
