Tesla’s pushing forward with its next-gen AI infrastructure. Elon Musk confirmed during a recent podcast that the company’s Cortex 2 GPU training cluster will be 500MW and reach operational status by mid-2026. The announcement came during a 3-hour conversation with Peter Diamandis, filmed at Giga Texas with a Cybercab positioned in the background.
Construction activity at the Texas facility indicates Tesla’s moving faster than anticipated on the project.
Drone footage analyst Joe Tegtmeyer has documented rapid progress on the cooling infrastructure required for Cortex 2. Two chiller plants currently under construction will feature at least 12 fan units, double the capacity installed for Cortex 1 on the southeast side of the main factory.
First plant is receiving FRP beam installations. Fiber-reinforced polymer structures form the housings and supports for cooling equipment. Meanwhile, the second plant is in its foundation phase, with rebar being installed to create exterior walls.
What’s changed? Tesla’s engineering team redesigned the layout for Cortex 2’s cooling system. Unlike the original Cortex 1 setup—where pumps, heat exchangers, manifolds, and valves sit at ground level, Cortex 2 houses much of this supporting equipment inside the structure itself. Modification streamlines operations and potentially reduces the physical footprint needed for Tesla’s GPU training cluster expansion.
Recent deliveries of FRP beams signal the aggressive timeline Tesla’s maintaining. Joe noted that beams delivered on January 5 were already installed by January 7. That’s a two-day turnaround from delivery to installation, a pace that suggests Tesla’s prioritizing this infrastructure build.
500MW requirement for Cortex 2 represents a substantial increase over existing facilities. Managing that energy load demands extensive cooling capacity, which explains the doubling of fan units and the redesigned chiller plant architecture.
Tesla needs this computational power for training its FSD neural networks and potentially other AI applications. Timeline Elon provided puts the facility online within months, making it one of the fastest large-scale AI infrastructure deployments in the industry.
Related Post
Tesla Cortex 2.0: Second $5B+ Computing Cluster Under Construction at Giga Texas
Tesla Cortex: The Supercomputer Powering the Future of AI and Autonomous Driving