Postdoctoral Researcher - OCCL (Open-source Accelerator Communication Library) ETH Zürich in Switzerland

The Singapore-ETH Centre, established by ETH Zürich and Singapore’s National Research Foundation, is a hub for research on urban sustainability, resilience, and health. This postdoctoral position is based at ETH Zürich, one of the world’s leading science and technology universities, and focuses on the development of a novel open-source accelerator communication library (OCCL) for distributed AI workloads.

The OCCL project addresses the growing computational and energy demands of large-scale artificial intelligence models. As AI models scale, efficient communication between accelerators and across network substrates becomes critical. The OCCL library will support both point-to-point and collective communication, initiated by hosts and device kernels, and will be compatible with a wide range of devices and high-performance network technologies such as NVLINK, InfiniBand, UltraEthernet, and UCX.

The postdoctoral researcher will co-lead the design and development of OCCL, focusing on communication primitives and algorithms optimized for machine learning training and inference. Responsibilities include investigating and optimizing communication patterns for deep learning workloads, ensuring portability and configurability across diverse hardware platforms, and evaluating performance on GPU clusters and large-scale computing systems. The role also involves collaboration with researchers and engineers at partner institutions in Singapore and Zurich, integration with the FastTrackAI software ecosystem, and co-supervision of MSc and BSc students.

Applicants must have a PhD in Computer Science or a related field, with strong expertise in parallel computing, compilers, performance optimization, and high-performance computer systems. Proficiency in C and C++ is required, along with experience in communication libraries (MPI, NCCL) and modern network technologies. Prior knowledge of communication optimization for large-scale ML workloads and experience managing larger software engineering projects are essential. Candidates should be able to work independently and communicate effectively with remote colleagues. English proficiency is mandatory.

ETH Zürich offers a diverse and inclusive workplace, comprehensive healthcare insurance, flexible hybrid work arrangements, annual leave, and networking opportunities. The Singapore-ETH Centre is committed to equal opportunity and family-friendly employment practices. Applications must be submitted online via the ETH Zürich portal, including a cover letter, CV with references, and university transcripts. Reference letters may be included. Email or postal applications will not be considered.

This position provides an excellent opportunity to contribute to cutting-edge research in distributed AI systems, high-performance computing, and software engineering, within a collaborative and international environment at ETH Zürich.