Anthropic Pays AI Researchers More Than Traditional Silicon Engineers for Chip Design
Anthropic is expanding its custom ASIC chip design team and offering salaries up to $850,000 for research engineers training AI models to design chips. In contrast, the traditional silicon engineers physically designing the chips are offered lower salaries ranging from $320,000 to $485,000. This salary disparity highlights the tech industry's high premium on AI automation over traditional hardware engineering. By training models like Claude to automate complex tasks like RTL generation and physical design, Anthropic aims to accelerate chip development cycles. The "Chip Design RL" role focuses on building reinforcement learning environments to teach AI models the chip design flow, including RTL code generation, verification, and PPA optimization. Despite the salary gap, both roles require highly overlapping skills, including experience with EDA tools, UVM, and physical chip development.
## BACKGROUND
Custom ASIC (Application-Specific Integrated Circuit) chips are tailored for specific tasks like AI training and inference to improve efficiency over general-purpose GPUs. Designing these chips traditionally involves complex steps like writing Register-Transfer Level (RTL) code and using Universal Verification Methodology (UVM) for testing, areas where tech companies are increasingly applying reinforcement learning (RL) to automate and optimize.