Senior Embedded Performance Engineer - Equity & Impact

Fractile; London Borough of Southwark; On-site; GBP 75,000 - 120,000; Full time

Benefits offered by this job

Equity

Private Medical

Dental and Vision

Contributory Pension

25 Days holiday + bank holidays

Life/Critical Illness Insurance

Job summary

Fractile is building the hardware and software stack to accelerate frontier AI workloads on chips and systems. The role focuses on writing and optimizing runtime software and ML kernels close to the hardware, influencing silicon and compiler decisions.

You will drive early accelerator compute runs, evaluate performance on silicon, run benchmarks, and feed results back into the roadmap with tight collaboration across teams.

Qualifications

• Proven deeply embedded software experience.
• Strong performance instincts.
• Ability to reason about low‑level architecture and memory behavior.

Responsibilities

• Write and optimise performance-critical ML kernels in C, with assembly where it matters.
• Build low-level control paths including scheduling, synchronization, and data movement.
• Write targeted validation workloads and microbenchmarks to keep simulation and hardware behaviour aligned.
• Profile, benchmark, and track regressions to ensure real performance gains.
• Collaborate with simulation, hardware, ML, compiler, firmware, and runtime engineers to translate profiling into architecture improvements.

Skills

C programming

Assembly (RISC-V)

Embedded software

Performance optimization

Low-level architecture knowledge

Debugging

Clear communication

Tools

RISC‑V ISA

Job description

Fractile is building the hardware and software stack to accelerate frontier AI workloads on chips and systems. The role focuses on writing and optimizing runtime software and ML kernels close to the hardware, influencing silicon and compiler decisions.

You will drive early accelerator compute runs, evaluate performance on silicon, run benchmarks, and feed results back into the roadmap with tight collaboration across teams.