Find your next role
Discover roles across ambitious teams
1,166 jobs
- More jobs
Mistral AI
Research Engineer, Gpu Performance
Paris, France · On-site · Full time · Mid level
About the role
Develop low-level GPU code to maximize performance of frontier AI models. Collaborate with scientists to optimize model architectures for efficient inference and scale.
What you’ll do
• Develop low-level code to maximize GPU capacity. • Rethink generative model architecture for efficiency. • Integrate code into MLOps framework. • Profile and benchmark large-scale GPU training. • Model LLM execution on various GPUs.
What you’ll bring
• Expert in custom low-level GPU kernels. • Knowledge of distributed GPU cluster infrastructure. • Understanding of generative AI and language models. • Strong software design and testing skills. • Self-starter, collaborative, low-ego.
Life at Mistral AI
Comprehensive benefits including healthcare, parental leave, retirement plans, relocation support, wellness programs, and allowances. Benefits vary by country.
About Mistral AI
Mistral provides full-stack AI solutions partnering with enterprises across industries to co-create customized AI systems for high-stakes problems.