Job description
Welcome to Codasip
We believe is the most innovative processor solutions company. We take pride in designing and developing cutting-edge, high-performance, and energy-efficient CPU cores from scratch, and our own automated proprietary tools to fully customize them. We give our customers a unique competitive advantage by empowering their system-on-chip developers to build the most innovative products.
Our processor cores are based on the RISC-V open architecture. The potential for customizing RISC-V is unlocked with the Codasip Custom Compute approach: our unique architecture description language, CodAL, and the powerful automated processor design tool, Codasip Studio. These are at the heart of our unique and groundbreaking RISC-V processor solutions.
Founded in 2014, we’ve grown into a thriving and talented global community. Ou r IP engineering team s work from offices spread across Europe, including our first and largest design center in the beautiful city of Brno, Czechia. Across Europe, we already have design teams in Cambridge, Bristol, Munich, Villeneuve-Loubet, Barcelona, Thessaloniki, Heraklion and Athens. The Codasip team is also based close to its customers, which means we have dedicated sales and application engineers in the USA, Japan, Korea, and China.
- Team: Codasip Labs
- Employment: Full-time, long-term cooperation
- Experience level: Mid-Senior
- Location of work: Design centers in Brno or Prague, the Czech Republic (hybrid mode)
ABOUT THE POSITION:
We’re looking for a Embedded SW Developer with a strong analytical mindset and hands-on experience in implementing and benchmarking algorithms on embedded processors. You’ll explore different cores and approaches to solve complex problems (e.g. AI inference, motor control, signal processing), and propose innovative solutions. The role includes close collaboration with hardware engineers to co-design and build demonstrators, including FPGA-based prototypes in both bare-metal and OS-based environments.
You will be part of the AI Team within Codasip Labs, dedicated to research, architect, design, and advancing capabilities on RISC-V processors across a range of application domains, including but not limited to AI. Your work will contribute to Codasip internal roadmap as well as several EU-funded research projects (e.g. NEUROKIT2E, ISOLDE, CHORYS, …).
YOUR ROLE AND MAIN FOCUS:
Research, implement, and optimize algorithms on a range of RISC-V embedded processors (e.g., low-power, high-performance)
Benchmark performance across cores and configurations, compare results with state-of-the-art solutions, and propose improvements to enhance efficiency, scalability, and robustness leveraging Codasip Studio technology
Explore and develop algorithms and tools for efficiently mapping neural networks onto Codasip’s target hardware architectures
Collaborate closely with hardware engineers to co-design and build prototypes and demonstrators
Support business development through technical presentations and demonstrators, including materials for customer engagements, conferences, application notes, and website articles
3–5 years of recent and relevant hands-on experience in embedded software development, with solid exposure to commercial systems and platforms including bare-metal environments (experience with OS-based environments is appreciated)
Strong programming skills in C/C++, with a good understanding of assembly language; familiarity with scripting languages (e.g., Bash, Python) is a plus
Experience in algorithm and low-level library development and optimization for embedded systems
Experience with neural network inference libraries, graph compilers, and performance tuning of AI/ML pipelines for embedded systems is appreciated
Familiarity with revision control systems (e.g. Git), continuous integration (CI), and software development best practices
Proven ability to collaborate across disciplines (e.g., working closely with hardware engineers, research teams, and external partners)
Good communication skills, experience with customer-facing activities and European co-funded projects is appreciated
SOME USEFUL LINKS ON CODASIP LABS:
- Codasip launches Codasip Labs to accelerate advanced technologies
- Codasip Studio RISC-V EDA tool - Codasip
- Codasip L150 and CORDIC acceleration for motor control
- NEUROKIT2E: A deep learning platform dedicated to embedded HW and Europe - Codasip
