4 of 4 Remote/Hybrid Performance Optimisation Jobs in Cambridge

practice is essential, including graph processing, operator execution, memory management, and backend integration. This role involves analysing, investigating, diagnosing, and resolving complex functional or performance issues across frameworks, runtimes, compilers, drivers, and hardware abstraction layers.The work brings together engineers across runtime, compiler, driver, and hardware teams to design, develop … optimise solutions within a large, modular codebase. Strong analytical and problem-solving skills, combined with an interest in machine learning systems, AI acceleration, and performance optimisation, will support success in the role.Beneficial but Not RequiredExperience across every area below is not expected. Relevant knowledge or experience could include ...

world while gaining meaningful experience across a variety of disciplines.About the teamAt Roku, the UI team builds the foundational frameworks, tooling, and high‐performance UI components that power experiences across millions of devices. We own the middleware and UI layers that enable smooth navigation, responsive interactions, rich animations … consistent developer workflows across the platform. Our work spans graphics rendering, layout and animation systems, performance optimization, and developer APIs that make it easy to create reliable, visually compelling applications at scale. We operate in a collaborative, high‐trust environment and partner closely with firmware, platform, and product teams ...

validated and deployable artifact on Arm-based systems.Our AI tooling will bring together model graph inspection, hardware compatibility analysis, graph partitioning, quantization and other performance optimization techniques with benchmarking and deployment packaging across Arm CPUs, GPUs, Ethos-U NPUs and will evolve alongside Arm's roadmap.We are looking … PyTorch, ONNX, ExecuTorch, Llama.cpp, vLLM and LiteRT.Understanding of model deployment to edge, embedded or heterogeneous computing, with an understanding of trade-offs between accuracy, performance and power consumption.You should also have:An engineering or computer science background.Experience taking a technical product from concept to adoption and an ability ...

deployable artifact on Arm-based systems. Our AI tooling will bring together model graph inspection, hardware compatibility analysis, graph partitioning, quantization and other performance optimization techniques with benchmarking and deployment packaging across Arm CPUs, GPUs, Ethos-U NPUs and will evolve alongside Arm's roadmap. We are looking … ONNX, ExecuTorch, Llama.cpp, vLLM and LiteRT. Understanding of model deployment to edge, embedded or heterogeneous computing, with an understanding of trade-offs between accuracy, performance and power consumption. You should also have: An engineering or computer science background. Experience taking a technical product from concept to adoption ...