Research Scientist Visual Generative AI & World Models

About Graphcore

At Graphcore, we’re building the future of AI compute. We’re a team of semiconductor, software and AI experts, with deep experience in creating the complete AI compute stack - from silicon and software to infrastructure at datacenter scale. As part of the SoftBank Group, backed by significant long-term investment, we are delivering key technology into the fast-growing SoftBank AI ecosystem. To meet the vast and exciting AI opportunity, Graphcore is expanding its teams around the world. We are bringing together the brightest minds to solve the toughest problems, in a place where everyone has the opportunity to make an impact on the company, our products and the future of artificial intelligence.

Job Summary

As a Research Scientist at Graphcore, you will advance AI research at the intersection of visual generative modelling, multimodal learning, world models and hardware-aware machine learning. You will explore new model architectures, training methods and deployment strategies with applications in embodied AI, robotics and autonomous systems. Example research directions could include efficient video generation, diffusion and flow-based models, multimodal representation learning, world models for agents, or analysis of how emerging generative AI workloads influence future AI accelerators. This role sits at the interface between frontier model research and AI hardware. Specialised hardware has been a key driver of AI progress over the last decade, and we believe that hardware-aware AI algorithms and AI-aware hardware developments will continue to be critical to advancing this field. We are looking for researchers and engineers with the theoretical depth, practical judgement and implementation skills to turn ambitious ideas into rigorous experiments, publications and technical insights that influence the future of AI compute.

The Team

Graphcore Research participates in both fundamental and applied research to characterise the computational requirements of machine intelligence and to demonstrate how hardware can drive the next generation of innovative AI models. We publish at leading AI/ML conferences, including NeurIPS, ICML and ICLR, as well as specialist workshops, and collaborate with other research teams and organisations across the world.

We pride ourselves on being a supportive and collaborative team, where we organise around individual research interests and solve problems together. Our work spans efficient compute, model scaling, distributed training and inference, and AI models for multiple modalities and applications, including sequence- and graph-based data. We’re based across London, Cambridge and Bristol, with projects and discussions that involve all our locations.

Perhaps the best way to get an idea of what we’re all about is to read one of our papers or an article on our blog . If you’re excited to work at the cutting edge o