Senior Software Engineer - AI Compute Libraries & Performance

📋 Description

• Build high-performance AI compute libraries for Graphcore’s hardware close to the hardware.
• Own kernels for linear algebra and tensor ops (GEMM, convolutions, reductions, fused ops).
• Improve performance, correctness and numerical reliability across AI workloads.
• Design and implement kernels in C++ with focus on performance and scalability.
• Profile, test, debug, and optimize software with clear ownership and fast feedback.
• Mentor colleagues, share knowledge and raise technical bar within the team.

🎯 Requirements

• Excellent programming and scripting skills using C++ and Python
• Design and optimize kernels for linear algebra and tensor operations
• Experience with BLAS, PyTorch or similar stacks; ability to extend kernels
• Understanding of processor architectures and profiling on Linux
• Strong focus on performance, numerics, and memory/layout optimizations
• Strong communication, teamwork and mentoring abilities

🎁 Benefits

• Unlimited annual leave
• Up to 5% matched pension
• Phantom equity – share in Graphcore’s success
• True flexibility in how and where you work
• Office spaces designed for collaboration
• Free food and an on-site barista