ETP4HPC Tackling Quantum: HPC-QC Software Stack Developments for Hybrid Systems and Applications

The technology to design and engineer quantum computers and simulators (QCS) is still maturing to achieve reliable and scalable QCS systems for practical applications. There are notable efforts underway in Europe and globally to interface QCS and HPC systems. A primary objective for this has been towards operating QCS systems as local or remote accelerators for HPC systems and applications. One of the main emergent mechanisms for provisioning access to QCS systems to end-users is HPC centres across the world interfacing classical HPC systems with QCS systems - either co-located within the same datacentre or remotely accessible from a HPC cluster as a network of federated systems. This requires extensions to classical HPC stack elements - particularly at the levels of schedulers, compilers, parallel programming models and languages, applications and workflows - to seamlessly target QCS systems as accelerators. However, the programming stack for QCS systems is predominantly composed of lower-level system software which is device- and/or vendor-oriented using higher-level software libraries for ease of programming, often delivered within Python, C++ or Julia frameworks. This leaves significant gaps, and opportunities, to further develop existing HPC software components at different levels of the stack to natively target QCS systems as accelerators from within the HPC software stack. This session brings together experts and audiences from HPC and QCS communities to focus on the full-stack to discuss, prioritise and identify technical development opportunities. The outcomes would inform future work programmes, supported by ETP4HPC, about the objectives and impacts on which to focus.

---

Contributors:

• Bruno Taketani
• Bettina Heim
• Jean-Philippe Nominé
• Hans-Christian Hoppe