Accelerated Computing Engineer

Aerospacelab; Mont-Saint-Guibert; On-site; EUR 70,000 - 110,000; Full time

Job summary

Aerospacelab in Belgium is seeking an engineer to turn processing algorithms into efficient code for embedded hardware accelerators on satellites.

You will port, profile and optimize algorithms to run in real time within tight power, memory and thermal limits, collaborating closely with algorithm developers, flight software and hardware engineers; your code will operate in orbit.

Qualifications

• BSc, MSc or PhD in Computer Engineering, Electrical Engineering, Applied Physics or related field.
• Professional experience programming hardware accelerators (e.g. CUDA, OpenCL, SYCL or HLS).
• Solid understanding of parallel architectures: memory hierarchy, data movement, pipelining and parallelism.
• Experience optimizing numerical code for performance (SIMD, multithreading, profiling).
• Background in digital signal processing and/or image processing.
• Experienced developing on Linux.
• Proficient in English.
• Eligible to work from one of our office locations (Belgium / Switzerland / France).

Responsibilities

• Port reference algorithms (Python/MATLAB) to optimized C++ implementations on hardware accelerators.
• Profile and optimize throughput, latency, memory bandwidth and power consumption.
• Work with hardware engineers on how processing is split across computer resources.
• Build validation tests that check optimized code against reference models.
• Integrate processing pipelines with the on-board software.
• Write technical documentation and take part in code reviews.

Skills

C++ development

CUDA/OpenCL/SYCL

Real-time processing

Embedded Linux

Parallel architectures

Code optimization

English proficient

Education

BSc/MSc/PhD in Computer Engineering, Electrical Engineering, Applied Physics or related field

Tools

CUDA

OpenCL

SYCL

HLS

Linux

Job description

About Aerospacelab

Our mission is supported by an ambitious vertically integrated approach: design & manufacture of small satellites combined with development of earth observation services.

Since its creation in 2018, the company has already grown substantially. Our team gathers more than 25 engineering expertise, from hardware design to software development & data science.

In 2026, Aerospacelab counts offices in Louvain-la-Neuve (Belgium), Toulouse (France), Lausanne (Switzerland) and US West Coast with more than 450full-time employees, with the ambition to position itself as the European leader in small satellites.

Role overview

Aerospacelab designs, builds and operates satellites, and more and more of ourdata processing is moving on board. We are looking for an engineer who can turnprocessing algorithms into efficient code on embedded hardware accelerators.

You will take algorithms from our processing teams, then port, profile andoptimize them so they run in real time within the power, memory and thermallimits of a satellite. You will work closely with algorithm developers, flightsoftware engineers and hardware engineers. Your code will run in orbit.

What you will work on

• High-throughputsignal processing pipelines.


• Imageprocessing chains, from raw sensor data to calibrated products.


• Real-timeprocessing of high-rate sensor data.


• Reducingdata volume on board before downlink.

Our tech stack

• Languages:C++, CUDA and other accelerator frameworks, Python/NumPy for referencemodels.


• Platform:embedded Linux on heterogeneous computing hardware.


• Tooling:GitLab CI, Docker, cmake, performance profilers.

Key activities

• Portreference algorithms (Python/MATLAB) to optimized C++ implementations onhardware accelerators.


• Profileand optimize throughput, latency, memory bandwidth and power consumption.


• Workwith hardware engineers on how processing is split across computeresources.


• Buildvalidation tests that check optimized code against reference models.


• Integrateprocessing pipelines with the on-board software.


• Writetechnical documentation and take part in code reviews.

Must have

• BSc,MSc or PhD in Computer Engineering, Electrical Engineering, AppliedPhysics or a related field.


• Professionalexperience programming hardware accelerators (e.g. CUDA, OpenCL, SYCL orHLS).


• Solidunderstanding of parallel architectures: memory hierarchy, data movement,pipelining and parallelism.


• Experienceoptimizing numerical code for performance (SIMD, multithreading,profiling).


• Backgroundin digital signal processing and/or image processing.


• Experiencedeveloping on Linux.


• Proficientin English.


• Eligibleto work from one of our office locations (Belgium / Switzerland / France).

Nice to have

• Experiencewith embedded accelerator platforms and power-constrained systems.


• Knowledgeof optimized math libraries (FFT, linear algebra).


• Experiencewith heterogeneous designs combining processors, accelerators and FPGAs.


• Experiencein the space sector or with radiation-tolerant computing.