About the job

OLIX is at the forefront of AI hardware innovation, addressing the massive infrastructure gap created by the rapid growth of AI. The company is building the Optical Tensor Processing Unit (OTPU), a new paradigm delivering performance and energy efficiency impossible with existing chips. The Staff Performance Modelling Engineer will create and own analytical and simulation models that guide OTPU architecture and software evolution, building functional simulators as well as high-fidelity, cycle-accurate models to explore design trade-offs and deliver actionable insights. Responsibilities include defining and delivering technical vision, collaborating with engineering teams to ensure models reflect reality, educating peers on modelling methodologies, designing functional and performance simulators, performing workload analysis on ML workloads (LLMs, diffusion, graph workloads), conducting design-space exploration, and developing Python/C++ tooling integrated into CI for automated performance testing. Candidates should have 7+ years experience in performance or power modelling for CPUs, GPUs, ASICs, or accelerators, proven technical leadership with 5–10 engineers, strong C++/Python skills, familiarity with cycle-accurate simulators (gem5, SystemC, custom), a strong grasp of computer architecture fundamentals, familiarity with ML frameworks (PyTorch, TensorFlow, JAX), comfort reading RTL or schematics, excellent data visualization and communication skills, and a relevant bachelor’s degree (advanced degree preferred). Compensation includes a competitive salary of £211,000+, meaningful equity, a £24k proximity bonus, premium healthcare, 25 days annual leave plus UK bank holidays, access to high-spec ergonomic workspace, elite hardware (M4 Macs with M4 Pro upgrades), rapid prototyping labs, chef-prepared meals, and full UK/international visa sponsorship. Candidates from certain restricted countries may not be eligible due to US export control regulations.