Graduate Software Engineer, Accelerator Modelling

Graphcore Early Careers; West of England; Vor Ort; GBP 42.000 - 55.000; Vollzeit

Benefits dieser Stelle

Pension match

Phantom equity

Private medical insurance

Dental cover

Zusammenfassung

Graphcore is offering a two-year Graduate Software Engineer programme in Accelerator Modelling in Bristol. You will develop, test and refine models and simulations for next‑generation AI accelerators, starting on day one with ownership and learning alongside a supportive team.

You will gain hands‑on experience in C++ development, hardware simulation and performance modelling, while contributing to planning, design and testing and seeing how modelling results inform hardware decisions across the

Qualifikationen

• Bachelor’s or Master’s degree in Computer Science, Maths, ML, DS or related field.
• Strong C++ programming skills demonstrated through study, projects or work.
• Understanding of CS fundamentals and interest in lower-level software.
• Ability to debug complex problems and explain your thinking clearly, with exposure to debugging tools, perf analysis, testing and collaboration.

Aufgaben

• Develop, test and refine models and simulations for next‑generation AI accelerators.
• Predict performance, power consumption and scalability for hardware decisions.
• Investigate workloads, contribute to planning, design, implementation and testing; help integrate future hardware versions.

Kenntnisse

C++ programming

Debugging skills

Performance modelling

Hardware awareness

Ausbildung

Bachelor's/Master’s degree in CS or related

Tools

Version control

Jobbeschreibung

Graphcore is offering a two-year Graduate Software Engineer programme in Accelerator Modelling in Bristol. You will develop, test and refine models and simulations for next‑generation AI accelerators, starting on day one with ownership and learning alongside a supportive team.

You will gain hands‑on experience in C++ development, hardware simulation and performance modelling, while contributing to planning, design and testing and seeing how modelling results inform hardware decisions across the