Note: The job is a remote job and is open to candidates in USA. Cobalt builds expert data and evaluation infrastructure for AI developers. The company is seeking senior hardware engineers and computer architects to design, implement, test, and refine command-line hardware engineering tasks that evaluate whether frontier AI models can perform real hardware design and verification work.
Responsibilities
- Design self-contained command line tasks drawn from real hardware engineering work, such as finding and fixing a functional bug in a Verilog or SystemVerilog design, writing a testbench that exposes a subtle corner case, repairing a design that fails timing or synthesis, proving or disproving a property with formal verification, debugging a pipeline hazard or cache coherence problem in a processor model, evaluating an architectural trade-off in a simulator, and bringing up bare-metal firmware on an emulated processor
- Build the task environment using open-source tools, write a reference solution, and write automated tests that verify whether a solution is correct, for example through simulation against a reference model
- Test your task against a frontier AI model and refine it until the model fails for substantive reasons rather than because of ambiguity or trick wording
- Respond to reviewer feedback until the task is accepted
Skills
- A PhD in electrical engineering, computer engineering, computer science, or a closely related field
- Academic or industry experience in computer architecture, digital design, design verification, or a closely related area of chip development
- At least one publication, either academic (for example a peer-reviewed paper at a recognized architecture or design automation venue) or professional (for example a conference talk, a granted patent, or a widely used open-source hardware project)
- Expert-level skill in Verilog or SystemVerilog, and working knowledge of C or C++ and Python
- Familiarity with open-source hardware tools, such as Verilator, Icarus Verilog, Yosys, gem5, or the RISC-V toolchain and QEMU. Experience with commercial tools transfers well, but tasks must be built with open-source tools only
- Fluency in the Linux command line, shell scripting, Docker, and Git
- The ability to write precise task specifications that another engineer could follow without asking questions
Benefits
- Expand your influence through evaluation projects, advisory roles, and research collaborations, while deepening your understanding of how frontier models are trained and assessed.
- Collaborate with researchers from leading institutions and labs on high-impact, flexible work.
- Flexible 10 to 40 hour weeks that fit around your research position and your life.
Company Overview