Tesla has completed the design of a new generation of its Artificial Intelligence chips for Full Self-Driving. Elon Musk said earlier that the AI5 chip will offer performance comparable to NVIDIA’s ‘Hopper’ architecture, with two AI5 units matching the power of a single ‘Blackwell’ processor. The new chip marks the next step in Tesla’s in-house compute development for its driving systems.
Tesla is splitting manufacturing of its new AI5 accelerator between Samsung and TSMC. The chips will be produced at Samsung’s plant in Taylor, Texas, and TSMC’s facility in Arizona. The manufacturing split is intended to diversify the supply chain and keep chip supply under control for different demand scenarios. For Samsung, the deal was described in late 2025 reports as a notable win for its foundry business.
The project also includes memory supplied by SK hynix. The DRAM appears to be LPDDR5X memory integrated into the package, with modules arranged on both the left and right sides. On both the left and right sides, there are two rows of SK hynix LPDDR5X memory modules, each with three modules. This totals 12 LPDDR5X memory modules per AI5 chip. With 16 GB per module, this results in 192 GB of LPDDR5X memory per single AI5 SoC.
