Pauline J. Ollitrault, Abhinav Kandala, et al.
PRResearch
Supercomputers and quantum computers both aim to revolutionize the computing world by tackling challenges well beyond the reach of previous processing power. These technologies have, for the most part, been evolving in separate ways, but is this the best approach? We argue that the combination of quantum computing and HPC (QHPC) is not only possible, but also inevitable. In this tutorial, we delve into the challenges we have faced while creating hybrid quantum-classical pipelines. By raising these challenges, we aim to educate and inform HPC and quantum engineers about the integration pitfalls and best practices for overcoming them, thereby making the QHPC field more accessible. We will explore the need and challenges associated with building pipelines that require low-latency communication between QPUs and CPUs. This will involve exploring use cases that require such closely coupled architectures, the different ways these architectures can be built, and the development of an example workflow that takes advantage of fast information exchange between QPUs and CPUs. We will then examine how to best leverage HPC to mitigate QPU overhead. This will include an in-depth examination of an algorithm that dynamically switches between simulation and hardware runs to approximate varying problem sizes during training. These are open-ended questions, and we will explore the different approaches behind them before focusing on our own experiences. This will be done through open dialogue, interactive polls and Jupyter notebook coding implementations. We hope attendees will come away feeling equipped to embark on their journey into QHPC.
Pauline J. Ollitrault, Abhinav Kandala, et al.
PRResearch
Elisa Bäumer, Vinay Tripathi, et al.
APS March Meeting 2024
Max Rossmannek, Fabijan Pavošević, et al.
ACS Fall 2023
Petar Jurcevic, Luke Govia
APS March Meeting 2023