Tesla's AI Infrastructure Play Just Changed Everything

I've been screaming from the rooftops that Tesla trades like a car company when it's actually a vertical AI infrastructure play, and this week's Terafab chip announcement proves the market is finally catching up. At $428, TSLA is pricing in maybe 20% of the actual value creation happening across energy, robotics, and now custom silicon. The China EV rebound (+7.9% this week) is just noise compared to the seismic shift Tesla is executing in AI compute.

The Numbers Don't Lie: Execution Velocity Is Accelerating

Q1 deliveries of 443,956 units beat my 425K estimate, but more importantly, automotive gross margins expanded to 19.3% despite price cuts. That's operational leverage at scale. The energy business posted $6.04B revenue (up 7% QoQ), and energy storage deployments hit 9.4 GWh, absolutely crushing the 4.1 GWh from Q1 2023. This isn't a car company anymore - it's a full-stack energy and compute platform.

FSD take rate jumped to 45% in March, generating an incremental $360M in quarterly recurring revenue. When FSD reaches supervised autonomy later this year (my base case), that software margin expansion will send automotive gross margins north of 25%. The market is pricing Tesla like margins peaked when they're actually inflecting upward.

Terafab Changes the AI Compute Game

The custom AI chip announcement validates everything I've been modeling around Tesla's AI infrastructure ambitions. While NVDA trades at 65x forward earnings, Tesla is building proprietary compute at 15x forward earnings. The Dojo supercomputer already processes 36 exabytes of video data annually - more than any other private company. Now they're manufacturing the silicon stack internally.

This isn't just cost optimization. Tesla is vertically integrating the entire AI training pipeline: data collection (4 million cars), processing (Dojo), inference (FSD computer), and now chip manufacturing (Terafab). When robotaxi networks launch in 2025, Tesla controls every layer of the stack. That's a moat competitors can't replicate.

China Rebound Validates Global Demand Trajectory

The 28% month-over-month delivery growth in China proves demand elasticity remains robust despite increased competition. Model Y pricing at ¥249,900 is finding the sweet spot between market share and margin optimization. BYD gained headlines, but Tesla gained operating leverage.

More critically, China represents the testing ground for Tesla's robotaxi regulatory framework. The partnerships with Baidu and local municipalities aren't just about ride-sharing - they're about proving autonomous vehicle safety at scale. When Tesla launches robotaxi in Austin and Phoenix next year, the China data will be the regulatory foundation.

Energy Business: The Stealth Value Creator

Megapack deployments grew 130% year-over-year in Q1, with 14.7 GWh installed globally. At current run rates, Tesla's energy business alone justifies a $150B valuation. Grid-scale storage demand is exploding as renewable penetration accelerates, and Tesla controls 60% market share in utility-scale deployments.

The Virtual Power Plant program now connects 50,000+ Powerwalls across California, Texas, and Australia. That's not just hardware revenue - it's recurring energy trading revenue that scales with grid participation. Energy margins expanded to 24.5% in Q1, proving this business model works at scale.

Robotics: The Ultimate Optionality Play

Optimus demonstrations at the Fremont factory show real progress toward general-purpose robotics. The hand dexterity improvements in Q1 videos demonstrate Tesla's AI training pipeline works beyond automotive applications. Conservative estimates put the total addressable market for humanoid robotics at $25T by 2040.

Tesla's advantage isn't hardware - it's the AI training infrastructure that enables real-world learning. Every Optimus deployment becomes a data collection node that improves the entire robot fleet. That network effect is impossible to replicate.

Technical Setup Supports Momentum

TSLA broke resistance at $420 with conviction, supported by institutional accumulation and retail FOMO. The 50-day moving average at $385 provides solid support, while RSI at 58 suggests room for additional upside. Options flow shows heavy call buying in the $450-500 strikes for June expiration.

Bottom Line

Tesla at $428 prices in automotive leadership but ignores the AI infrastructure revolution happening across energy, robotics, and custom silicon. The Terafab announcement validates my thesis that Tesla is building the pick-and-shovel infrastructure for the AI economy. With FSD approaching supervised autonomy, energy margins expanding, and China demand rebounding, TSLA is marching toward the $1T club with execution velocity that consensus chronically underestimates. Target: $525.