News
DealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory’s NVIDIA & AMD Enterprise ExpansionDealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory’s NVIDIA & AMD Enterprise Expansion
EveryNeocloud
Cortex

Cortex

Tesla's AI training cluster for autonomous driving and robotics.

Cortex is a supercomputer cluster built by Tesla in Austin, Texas. It uses over 100,000 Nvidia H100 and H200 chips to train artificial intelligence for self-driving cars and humanoid robots. The system supports Tesla's shift toward autonomy and robotics.

Share on WhatsApp

  • Located in Austin, Texas
  • Uses over 100,000 Nvidia H100/H200 chips
  • Supports Full Self-Driving and Optimus
  • Future power demand of 500 megawatts
  • Part of Tesla's AI pivot

What it is

Cortex is a supercomputer cluster developed by Tesla. It is located at the company’s headquarters in Austin, Texas. The system is designed to process vast amounts of video data collected from Tesla vehicles worldwide. This data trains the Full Self-Driving software and the Optimus humanoid robot. The cluster relies on more than 100,000 Nvidia H100 and H200 graphics processing units. These chips handle the heavy computational load required for video-based AI training. Cortex represents a key part of Tesla’s strategy to become a leader in autonomy and robotics rather than just electric vehicles.

How it came about

Tesla has been building large data centres at its Gigafactories to support its AI ambitions. While the Dojo supercomputer is being developed in New York, Cortex is situated in Texas. Another operational cluster exists in Memphis and is currently undergoing expansion. Elon Musk stated that the Memphis facility aims to be the most powerful AI-training cluster globally. The development of Cortex coincides with Tesla’s pivot toward robotaxis and humanoid robots. This shift comes as the company faces slowing sales in its traditional electric vehicle market. The focus on AI infrastructure reflects the high cost of competing in this sector, which Musk noted requires billions in annual spending.

The compute

The Cortex cluster runs on over 100,000 Nvidia H100 and H200 GPUs. These processors are essential for training neural networks using video footage. The hardware is housed in black server racks with thick bundles of red and blue cables connecting them. The system generates significant heat, requiring a robust cooling solution. Tesla uses a patented cooling system with large fans on the roof, described as being similar in size to aircraft propellers. The loud hum from these overhead systems indicates the intense energy usage. This infrastructure allows Tesla to process data from millions of cars to improve its autonomous driving algorithms.

Where the GPUs are

Cortex is based in Austin, Texas, on the south side of Tesla’s headquarters. The facility resembles other large data centres but is tailored for AI workloads. In the future, the Texas supercomputer is expected to require 500 megawatts of power. To put this in perspective, a single sports stadium might use 5 megawatts during a game. This means Cortex could consume as much electricity as 100 stadiums running simultaneously. This high power demand highlights the energy-intensive nature of modern AI training clusters. Tesla also operates other AI facilities in New York and Memphis.

Its place in AI

Cortex plays a central role in Tesla’s transition to an AI and robotics company. It supports the development of Full Self-Driving technology and the Optimus robot. Elon Musk has claimed that the associated Memphis cluster will be the world’s most powerful AI-training system. This infrastructure is critical for processing the colossal amounts of video data needed for end-to-end neural network training. By building these capabilities in-house, Tesla aims to maintain a competitive edge in autonomy. The scale of investment underscores the importance of compute power in the current AI landscape.

Questions and criticism

Tesla’s Full Self-Driving technology has shown flaws in real-world conditions and faces regulatory scrutiny. The company is under investigation by the U.S. Department of Justice regarding how it markets this technology. There are unresolved questions about liability in the event of a robotaxi crash. Additionally, Musk’s decision to prioritise sending Nvidia chips to his social media platform X raised concerns. Critics note that X is privately held and technically separate from Tesla. The massive power consumption of facilities like Cortex also raises questions about energy sustainability.

Questions people ask about Cortex

What is the Cortex supercomputer?

Cortex is a supercomputer cluster built by Tesla in Austin, Texas. It uses over 100,000 Nvidia H100 and H200 chips to train AI models for autonomous driving and humanoid robots.

How much power does Cortex use?

The facility is expected to require 500 megawatts of power in the future. This is equivalent to the energy usage of about 100 sports stadiums during a game.

What is Cortex used for?

It processes video footage from Tesla vehicles to train the Full Self-Driving software. It also supports the development of the Optimus humanoid robot through AI training.

Good reading and sources

    Where this page comes from

    There is no English Wikipedia article on Cortex yet, so the facts on this page come from the sources listed above: tomshardware.com/desktops/servers/elon-musk-shows-off-cortex-ai-superc, insideevs.com/news/731363/tesla-cortex-supercomputer-giga-texas/, convergedigest.com/tesla-activates-cortex-2-data-center-at-gigatexas/. Some are the subject's own pages, so its claims about itself are reported as its claims. EveryNeocloud wrote the text in its own words, with AI assistance, and compares it with the sources to make sure no passage is copied.

    It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.

    Photo credits on this page

    The small pictures above come from Wikimedia Commons under free licences. Some are cropped to fit.

    The weekly digest

    Every Monday: the provider of the day, new pages on EveryNeocloud, and the conferences and anniversaries of the week ahead.