Senior Embedded DSP & AI Optimization Engineer

Arm Limited; Cambridge; On-site; GBP 80,000 - 110,000; Full time

Job summary

Arm Limited in Cambridge seeks an experienced embedded software engineer to design, optimize, and present DSP and AI workloads on Arm platforms. You will collaborate with hardware and software teams to bring up platforms, port OSes, and demonstrate performance improvements on next-generation solutions.

The role combines low-level embedded development, processor architecture knowledge, and system-level optimization to deliver efficient compute pipelines across DSP, ML, and vision workloads.

Qualifications

• Strong proficiency in C/C++ and low-level embedded programming.
• Solid understanding of processor architectures, memory hierarchies, caches, DMA, and interrupt handling.
• Familiarity with RTOS, Zephyr, or Linux and embedded operating systems.

Responsibilities

• Platform bring-up and embedded platform configuration (bare-metal, RTOS, Zephyr, embedded Linux).
• Performance optimization of DSP, audio, vision and AI workloads using Arm features (MVE, NEON, SVE2, SME).
• Develop and optimize kernels and libraries for signal processing and ML preprocessing.
• System integration of sensors, memory, DMA, accelerators and software components into pipelines.
• Benchmarking and profiling CPU, memory, bandwidth, latency and power with hardware counters.

Skills

C/C++

Embedded programming

Processor architectures

RTOS/Linux

Performance profiling

DSP algorithms

Hardware boards

Collaboration

Communication

Tools

CMSIS-DSP

CMSIS-NN

Ethos-U NPU

Job description

Arm Limited in Cambridge seeks an experienced embedded software engineer to design, optimize, and present DSP and AI workloads on Arm platforms. You will collaborate with hardware and software teams to bring up platforms, port OSes, and demonstrate performance improvements on next-generation solutions.

The role combines low-level embedded development, processor architecture knowledge, and system-level optimization to deliver efficient compute pipelines across DSP, ML, and vision workloads.