Graphcore

Lead the architecture and development of shared software frameworks that automate testing and validation for Graphcore’s AI hardware platforms. This Principal-level role combines hands-on engineering, cross-functional technical leadership, and complex hardware-software debugging.

Develop and maintain high-performance C++ libraries and runtime APIs that enable engineers to build and execute workloads on Graphcore AI hardware. You’ll own library changes end to end while improving performance, reliability, documentation, and developer adoption.

Lead the strategy and implementation of system-level tests that protect Graphcore’s firmware, BMC, and Linux device-driver releases. This hands-on staff role combines low-level software expertise, test automation, reusable tooling, and technical leadership across complex server systems.

Build and optimize framework support that enables AI developers to use Graphcore accelerators with Triton and other leading ML frameworks. This role combines compiler development, performance engineering, open-source collaboration, and cross-team technical ownership.

Join Graphcore’s 12-week summer internship to build firmware that boots, monitors, and supports next-generation AI hardware. You’ll gain hands-on experience in embedded software, hardware debugging, testing, and collaborative development.

Join Graphcore’s permanent Graduate Software Engineer programme in Bristol and build analysis tools that help developers understand and optimise software running on Graphcore hardware. You’ll work across libraries, command-line tools and graphical applications while developing your engineering skills over a two-year programme.

Join Graphcore for a 12-week summer internship developing high-performance C++ compute libraries and kernels that help machine learning models run efficiently on next-generation AI hardware. You’ll gain hands-on experience in performance analysis, numerical computing, profiling, and software reliability.

Join Graphcore’s permanent Graduate Software Engineer programme to build high-performance C++ compute libraries and kernels that help AI models run efficiently on next-generation hardware. You’ll gain two years of hands-on experience in performance analysis, numerical validation, and low-level software engineering.