About the job
Join Graphcore as a Research Scientist
At Graphcore, we are at the forefront of AI compute innovation. Our team, comprised of semiconductor, software, and AI experts, is dedicated to developing an integrated AI compute stack, spanning from silicon and software to datacenter-scale infrastructure. As part of the SoftBank Group, we enjoy substantial long-term investment to push the boundaries of AI technology within the rapidly expanding SoftBank AI ecosystem. To harness the immense potential of AI, we are actively growing our global teams and inviting the brightest minds to tackle the most challenging problems, providing everyone the chance to make a significant impact on our company, products, and the future of artificial intelligence.
Role Overview
As a Research Scientist at Graphcore, your contributions will drive advancements in AI research by exploring innovative concepts that address critical AI/ML challenges. Over the past decade, specialized hardware has been pivotal in AI progress, and we believe that the synergy between hardware-aware AI algorithms and AI-aware hardware will be essential for continuing breakthroughs in this fascinating field. We seek passionate scientists and engineers equipped with the theoretical and practical expertise necessary for impactful AI research.
Ideal candidates will have experience in low-power, edge, and embodied AI applications, including robotics, autonomous driving, and augmented/virtual reality. Your work will involve training and deploying multimodal AI models in these contexts, focusing on areas like world models, real-time computer vision, and generating and reasoning over audio/video streams.
About Our Team
Graphcore Research engages in both fundamental and applied research, aiming to characterize the computational needs of machine intelligence and showcase how hardware can propel the next generation of innovative AI models. We regularly publish our findings at top AI/ML conferences such as NeurIPS, ICML, and ICLR, and collaborate with various research teams and organizations worldwide.
We take pride in our supportive and collaborative environment, organizing ourselves around individual research interests to collectively solve problems in areas like efficient compute, model scaling, and distributed training and inference of AI models for diverse modalities and applications, including sequence- and graph-based data. Our teams are spread across London, Cambridge, and Bristol, fostering projects and discussions that connect all our locations.
To get a deeper insight into our work, we encourage you to read one of our publications or explore an article on our website.
