About the job

Flux Computing is seeking a Performance Modelling Engineer to create and own analytical and simulation models for OTPU hardware, including functional simulators and high-fidelity, cycle-accurate models. The role involves designing functional and performance simulators, instrumenting machine learning workloads to identify throughput, latency, and utilization bottlenecks, conducting large-scale design-space exploration to guide software, hardware, and optical teams, developing Python/C++ tooling for trace parsing, statistical analysis, and visualization, and integrating models into CI pipelines to ensure performance validation with every RTL commit. Candidates should have 3+ years of experience building performance or power models for CPUs, GPUs, ASICs, or accelerators, strong C++ and Python coding skills, familiarity with discrete-event or cycle-accurate simulators such as gem5 or SystemC, a solid understanding of computer-architecture fundamentals including memory systems, interconnects, and queuing theory, familiarity with machine-learning workloads and frameworks like PyTorch, TensorFlow, or JAX, the ability to read RTL or schematics and discuss micro-architectural trade-offs, and excellent data visualization and communication skills to turn millions of simulation samples into actionable insights. Flux offers visa sponsorship, full relocation support, a competitive salary with stock options, BUPA healthcare, ergonomic setups, office perks, 25 days paid holiday plus UK bank holidays, and other benefits to support engineers in a fast-paced AI hardware and software environment.