Tesla has signed a landmark $16.5 billion contract with Samsung Electronics to manufacture its next-generation artificial intelligence (AI) chips, marking one of the most significant public chip fabrication deals by any automaker to date. Announced on July 22, 2025, the multi-year agreement confirms that Samsung will fabricate Tesla’s advanced AI chips, designed for autonomous driving and robotics at its Taylor, Texas foundry, using its most advanced 2nm-class process node.
This is a major milestone not only for Tesla’s AI and autonomy ambitions, but also for Samsung’s positioning in the logic foundry space, where it is seeking to compete more directly with global leader TSMC. The Taylor facility was built with support from the U.S. CHIPS and Science Act, which encourages domestic semiconductor production critical to national infrastructure, including automotive and AI workloads.
Strengthening the Tesla–Samsung Partnership
This deal is not Tesla’s first with Samsung. The two companies have been collaborating for years, starting with Tesla’s Hardware 3 chip, produced by Samsung on its 14nm node. Tesla’s Hardware 4, launched with the Model S and X refresh in 2023, was also reportedly manufactured by Samsung, this time on a more advanced 7nm process. These chips power Tesla’s Full Self-Driving (FSD) suite, enabling perception, planning, and vehicle control based on neural network inference.
The new deal to produce Tesla’s upcoming AI chips, internally believed to be successors to its current FSD and AI silicon, is a continuation of this trusted foundry relationship. Production is expected to begin in 2026 and scale significantly through the decade.
Dojo Chips and TSMC: A Parallel Supply Chain
While Samsung handles Tesla’s vehicle inference chips, TSMC—the world’s leading foundry—has produced chips for Tesla’s Dojo AI training supercomputer. Tesla’s D1 chip, a custom design optimized for matrix operations and neural network training, is believed to be manufactured on TSMC’s 7nm process. This diversification highlights Tesla’s strategy to leverage multiple foundries for different workloads: Samsung for edge inference (in-car chips), and TSMC for data center training workloads.
Tesla’s increasing investments in custom silicon underscore its vision for end-to-end control over its AI stack, from training to deployment. The company’s AI Day events and development of the Optimus humanoid robot further reinforce its long-term commitment to building powerful, energy-efficient chips tailored to its vertical use cases.
Why the Samsung Deal Is a Big Win
For Samsung Foundry, Tesla’s $16.5 billion commitment is a breakthrough moment. The company has been striving to win more high-profile clients in the AI and high-performance computing (HPC) sectors, where TSMC dominates with clients like Nvidia, AMD, and Apple. The Tesla contract will likely help Samsung validate its SF2 (2nm-class) process for commercial-scale production, particularly as the Taylor site becomes operational.
Samsung’s Taylor plant—expected to be fully online by 2026—is part of a broader U.S. investment strategy that includes over $40 billion in semiconductor infrastructure. Tesla’s choice to manufacture in Texas strengthens efforts to onshore critical semiconductor supply chains, reducing dependence on East Asia amid growing geopolitical tensions.
Tesla’s silicon strategy is part of a broader industry trend in which hyperscalers and tech giants are investing in custom chips for AI workloads. Companies like Amazon (Trainium, Inferentia), Google (TPU), and Meta have already developed in-house chips, and Tesla’s approach closely mirrors that of Apple, which designs its own silicon and tightly integrates hardware with software.
For investors, this deal reinforces two key trends: Tesla’s deepening vertical integration into AI hardware, and Samsung’s expanding relevance in logic foundry services. The commitment to manufacture in the U.S. also aligns with investor priorities around supply chain security, geopolitical risk management, and CHIPS Act-enabled growth.
Moreover, the size and duration of the Tesla–Samsung contract bring long-term revenue visibility to Samsung’s foundry division, which has historically lagged behind TSMC in capturing high-end logic orders.
Discover more from Semiconductors Insight
Subscribe to get the latest posts sent to your email.
