Our client has developed the next-generation of Tensor Processing Units. Custom-designed AI accelerators, which are optimised for training & inference of large AI models. With working prototypes already, they are now looking to scale the business as quickly as possible.

We’re seeking a highly experienced & motivated individual to design & build the software architecture for our next-generation GPUs. This role demands deep expertise in C & C++ programming, low-level programming, compiler construction & optimisation techniques.

Role:Staff Software Engineer

Location:London

Salary:DOE / Competitive + benefits

Responsibilities

• Design & develop the software architecture for the next-generation TPU, ensuring high performance & scalability.
• Collaborate with hardware engineers to integrate software & hardware components seamlessly.
• Optimise software performance through advanced techniques in low-level programming & compiler design.
• Develop & maintain machine learning frameworks & tools to leverage the full potential of the TPU.
• Conduct code reviews, provide technical mentorship, & guide other team members in best practices.
• Stay current with industry trends & advancements in GPU technologies, machine learning, & optical computing.
• Lead & participate in the development of technical documentation & specifications.
• Drive innovation & contribute to the strategic direction of the software engineering team.

Skills & Experience

• 8+ years of experience in software engineering with a focus on C & C++ programming.
• Extensive experience in compilers, low-level programming, & optimisation techniques.
• Proven expertise in machine learning & its applications in high-performance computing.
• Strong problem-solving skills & the ability to think critically & creatively.
• Experience in high-pace, dynamic work environments.
• Bachelor's degree in computer science, electrical engineering, telecoms engineering, mathematics, or a related field.
• Excellent teamwork & communication skills, with the ability to collaborate effectively with cross-functional teams.
• Personal projects are a key differentiating factor & hold more weight than other requirements.