Senior AI Compute Platform Engineer

Arm Limited; Manchester; Hybrid; GBP 70,000 - 110,000; Full time

Benefits offered by this job

Stock awards

Hybrid working

Generous leave policy

Job summary

Arm Limited in Manchester is seeking experienced engineers to join a team building the compute platform for AI at scale. You will contribute to AI workloads, optimization across Arm Neoverse platforms, and cloud-server technologies.

The role involves collaboration with open-source communities and internal teams, in a hybrid office setting in Manchester. We welcome candidates who can deliver complex technical work with autonomy, learn quickly, and engage across stack layers from low-level

Qualifications

• Strong software engineering and programming skills with C++ and Python.
• Willingness to explore unfamiliar software stacks and codebases.
• Familiarity with Linux development environments, profiling, debugging and performance analysis.
• Interest in end-to-end application and system performance across the stack.
• Experience delivering sophisticated technical work involving multiple contributors.
• Open-source contributions and collaboration experience.
• Experience optimizing software for Arm servers (AArch64) including SIMD technologies.

Skills

C++

Python

Unfamiliar stacks

Linux environments

Performance analysis

End-to-end performance

Delivery of complex work

Open-source collaboration

AArch64/Arm optimization

Education

Computational or numerate subject degree

Tools

ML frameworks & LLM inference

Containers & cloud/server infra

Batching/scheduling/KV-cache

Neon/SVE/SME SIMD on Arm

Job description

Arm Limited in Manchester is seeking experienced engineers to join a team building the compute platform for AI at scale. You will contribute to AI workloads, optimization across Arm Neoverse platforms, and cloud-server technologies.

The role involves collaboration with open-source communities and internal teams, in a hybrid office setting in Manchester. We welcome candidates who can deliver complex technical work with autonomy, learn quickly, and engage across stack layers from low-level