AI Kernel Engineer

Overview

Senior AI Kernel Developer — Remote (Europe) | Permanent

Join a well-funded AI semiconductor company developing next-generation accelerator hardware and the low-level software that unlocks its performance. You'll be working at the lowest levels of the software stack, building highly optimised compute kernels for custom AI hardware and collaborating closely with compiler, runtime and hardware engineering teams.

• Develop and optimise high-performance AI kernel libraries
• Write highly optimised low-level code for AI accelerators
• Optimise compute-intensive workloads for maximum throughput and efficiency
• Collaborate with compiler, runtime and hardware engineers
• Analyse performance bottlenecks and implement low-level optimisations
• Contribute to software enabling next-generation AI hardware

• Strong C/C++
• Experience developing high-performance kernel libraries
• Assembly / assembler programming
• SIMD, vectorisation or low-level optimisation experience
• Strong understanding of computer architecture and memory hierarchy
• Experience optimising performance-critical software

• MLIR
• LLVM
• AI compilers
• CUDA, Triton or GPU programming
• AI inference optimisation
• Embedded or accelerator software development

If you enjoy writing highly optimised low-level code and want to help shape the software stack for next-generation AI hardware, I'd be happy to tell you more. Please contact Harry Hansford for more information.