Senior AI Compute Kernels Engineer — High-Performance

Graphcore; West of England; Hybrid; GBP 110,000 - 170,000; Full time

Benefits offered by this job

Work-life balance

Private Medical Insurance

Pension

Income protection

Progression

Wellbeing

Job summary

Graphcore is seeking a Senior Software Engineer to build high-performance AI compute libraries for next-generation hardware. You will design and implement kernels for linear algebra and tensor operations in C++, optimize performance, and ensure numerical correctness while collaborating within an Agile team.

You will also profile across architectures, contribute microbenchmarks, and mentor peers, with a strong emphasis on code quality, reliability, and integration with PyTorch and related stacks.

Qualifications

• Strong C++ and Python programming skills.
• Experience with kernel development and performance profiling.
• Familiarity with BLAS/DNN stacks and CPU micro-optimisations.
• Experience integrating native code into PyTorch or similar.
• Understanding of processor architectures and Linux profiling.

Responsibilities

• Design and implement kernels for linear algebra and tensor ops (GEMM, batched GEMM, convolutions, reductions, elementwise and fused operations) in C++.
• Own performance and correctness – add microbenchmarks, regression tests, numerics validation.
• Profile and optimise for next generation of AI hardware – threading, cache locality, memory layout, and kernel launch efficiency.
• Debug issues, resolve bugs and improve the quality and functionality of the product.
• Actively engage in and support Agile ways of working within the team.
• Mentor colleagues within the team, sharing knowledge and providing guidance.

Skills

C++

Python

Kernel development

Perf optimization

GEMM ops

Threading

Vectorisation

PyTorch integration

Linux profiling

Tools

PyTorch

cuDNN

Job description

Graphcore is seeking a Senior Software Engineer to build high-performance AI compute libraries for next-generation hardware. You will design and implement kernels for linear algebra and tensor operations in C++, optimize performance, and ensure numerical correctness while collaborating within an Agile team.

You will also profile across architectures, contribute microbenchmarks, and mentor peers, with a strong emphasis on code quality, reliability, and integration with PyTorch and related stacks.