Chip Design Expert (RTL Design & Verification)

About the Role

Mercor is partnering with a leading AI research laboratory to create specialized chip design benchmarks aimed at training and assessing AI models. In this expert role, you will be instrumental in maintaining the integrity and quality of hardware engineering tasks. You will work on practical assignments that include RTL block design and repair, as well as the creation of thorough testbenches, supported by existing specifications and reference materials. Your work will directly shape the methodology for testing advanced AI against industry-grade hardware challenges, providing a critical evaluation of task viability and difficulty for a growing benchmark library.

Requirements

• The ideal candidate possesses at least five years of professional experience in RTL design or verification within the semiconductor, FPGA, or EDA sectors.
• You must demonstrate strong proficiency in Verilog and SystemVerilog, alongside hands-on experience with simulation tools and the creation of robust testbenches.
• The ability to read and analyze Python test code is necessary to evaluate the provided technical tasks effectively.
• Furthermore, you should be a meticulous and skeptical reviewer capable of articulating written feedback regarding flaws, design ambiguities, or difficulty levels.
• Familiarity with open-source tools such as Icarus Verilog, Verilator, Yosys, or cocotb, as well as experience with formal verification or AI benchmarking, is considered a significant advantage.

Responsibilities

• You will be responsible for reviewing tasks from start to finish, ensuring that all specifications, starter files, reference solutions, and automated tests align correctly.
• It is essential to confirm that every problem is fair and solvable, verifying that the reference solution passes and that all test criteria are derived clearly from the provided specifications.
• You will proactively identify flaws, such as ambiguous instructions, broken test logic, or leaked answers.
• Additionally, you will judge the complexity of each task to determine if it sufficiently challenges state-of-the-art AI models.
• Your role also involves writing clear, actionable feedback and suggesting specific fixes to improve the quality and difficulty of the benchmark library.

Benefits & Perks

• This position offers a fully remote and flexible work environment requiring approximately 15 to 20 hours of work per week.
• The compensation for this role is set at $100 per hour, with the opportunity to begin work immediately.