Job Description
About The Role
Tesla's AI Compute electrical design team builds the system boards that power autonomous driving, humanoid robotics, and the data centers that train and serve our models. You will own complex high-speed motherboards built around Tesla's custom neural-engine silicon and leading x86/Arm processors, balancing inferencing performance, efficiency, cost, and reliability — from a single vehicle ECU up to scale-up and scale-out clusters in the AI inference data center.
Responsibilities:
Own hardware end-to-end: system architecture, electrical specification, SPICE simulation, schematics, prototype build, bring-up, and validationTranslate open-ended electrical discussions into concrete circuit schematicsCollaborate cross-functionally with silicon design, post-silicon, PCB layout, firmware, mechanical, signal and power integrity, thermal, reliability, EMC, and program managementAuthor bring-up and validation plans, then verify every feature and interface across PVT cornersDocument electrical design requirements, calculations, and power budgets to a level that supports audit and reuseDesign for manufacturability, testability, and EMC compliance
Required Qualifications
Degree in Electrical / Computer Engineering, or equivalent industry experienceDemonstrated end-to-end PCB assembly design ownership, including hands-on bring-up, electrical validation, failure analysis, and system testStrong high-speed board design (>10 Gbps) with PCIe, MIPI, Ethernet, DisplayPort, USB, or HDMIComplex motherboard experience with high-performance x86/Arm processors and/or custom siliconWorking knowledge of AI SoC specifications — compute, memory bandwidth, network ingress/egress, peak power, storage — and benchtop bring-upHardware and software co-design experience optimizing inference for large language modelsHigh-power, high-efficiency voltage regulator designSolid analog and power fundamentals (MOSFETs, op-amps, switching regulators, gate drivers)Experience shipping high-volume products from concept to mass productionComfortable iterating with mechanical, thermal, and EMC partners under tight, evolving constraintsTrack record of driving suppliers to meet reliability, performance, and scalability targets across the design-to-production lifecycleProficient with bench tools — oscilloscopes, BERT, network analyzers, signal generators
Preferred
Board layout experience in Cadence Allegro or Altium toolsAI training and inference processor internals (CPU / NPU / memory controller)Inter-chip communication for scale-up AI clusters: 112G / 224G SerDes, UALink, scale-up EthernetCutting-edge memory: GDDR6/7, HBM, DDR5, LPDDR4/5/6