European Systems and Software Summit (EQS3): Four Years, Four Themes and Twelve Insights
Wednesday, June 24, 2026 3:45 PM to 5:15 PM · 1 hr. 30 min. (Europe/Berlin)
Foyer D-G - 2nd Floor
Project Poster
Integration of Quantum Computing and HPC
Information
Poster is on display.
Quantum computing is increasingly positioned as a future accelerator within high-performance computing (HPC) environments, yet today’s systems remain fragile, heterogeneous, and difficult to operate as reliable, shared infrastructure. As the HPCQC landscape expands, the community needs a common, practice-informed understanding of what “integration” should mean across software, operations, and user-facing service delivery beyond current vendor-specific demonstrations and isolated pilots.
This poster distills consensus findings from six European Quantum Systems and Software Summit (EQS3) community sessions held since 2022 which brought together experts from supercomputing centers, quantum hardware providers, software developers, and application teams to identify recurring conceptual and technical bottlenecks that limit time-to-solution and slow scientific discovery. We synthesize here these observations into a framework of twelve insights within four themes that accepts near-term operational realities while defining mid-term architectural and programmatic choices for quantum-accelerated supercomputing:
Theme 1, HPC-first operational foundations, frames the QPU as a governed accelerator inside an HPC computing environment. Scalable hybrid HPCQC depends on HPC-native identity, policy, accounting, and scheduling as the system of record with quantum functions integrated via adapters rather than parallel control planes. It also treats multi-modality as permanent, implying that convergence is realistic only at carefully chosen layers where workflows can align and lower layers must remain modality-aware.
Theme 2, interfaces and ecosystem governance, highlights that interfaces rapidly become de facto standard. Therefore, they must stabilize meaning (e.g., job states, availability, results) to avoid “assumption lock-in” while also creating durable, well-scoped hook points for ISV contribution and clear boundaries between open stacks and vendor backends.
Theme 3, scheduling, allocation, and lifecycle operations, argues that time-to-solution is usually dominated by orchestration and queueing and not by quantum execution. The mismatch in time scales, from minutes to hours for scheduling versus milliseconds to seconds for runs, makes scheduler integration a non-optional architectural issue. It also notes that no single allocation model wins. System design should support multiple hybrid modes (co-scheduling, interleaving, malleability, time-multiplexing) and clearly separate device execution semantics from scheduler lifecycle control (acquire/release, accounting, policy).
Theme 4, reliability, portability, and evidence-driven convergence, treats drift, calibration cycles, and partial failures as first-class design constraints which needs standardized methods like nuanced health states (beyond up/down), telemetry, and logging as well as portability across multiple IRs and access modes without fragmenting semantics or provenance.
Aligned with the “HPC Connects” theme, this work connects stakeholders across facilities, vendors, and user communities to converge on pragmatic integration strategies. The poster provides a compact set of points and open questions realized in the EQS3 sessions. Past and future sessions intend to accelerate coordinated roadmap planning, reduce duplicated effort, and improve the credibility of HPCQC as a production-level capability over the years to come.
Contributors:
Quantum computing is increasingly positioned as a future accelerator within high-performance computing (HPC) environments, yet today’s systems remain fragile, heterogeneous, and difficult to operate as reliable, shared infrastructure. As the HPCQC landscape expands, the community needs a common, practice-informed understanding of what “integration” should mean across software, operations, and user-facing service delivery beyond current vendor-specific demonstrations and isolated pilots.
This poster distills consensus findings from six European Quantum Systems and Software Summit (EQS3) community sessions held since 2022 which brought together experts from supercomputing centers, quantum hardware providers, software developers, and application teams to identify recurring conceptual and technical bottlenecks that limit time-to-solution and slow scientific discovery. We synthesize here these observations into a framework of twelve insights within four themes that accepts near-term operational realities while defining mid-term architectural and programmatic choices for quantum-accelerated supercomputing:
Theme 1, HPC-first operational foundations, frames the QPU as a governed accelerator inside an HPC computing environment. Scalable hybrid HPCQC depends on HPC-native identity, policy, accounting, and scheduling as the system of record with quantum functions integrated via adapters rather than parallel control planes. It also treats multi-modality as permanent, implying that convergence is realistic only at carefully chosen layers where workflows can align and lower layers must remain modality-aware.
Theme 2, interfaces and ecosystem governance, highlights that interfaces rapidly become de facto standard. Therefore, they must stabilize meaning (e.g., job states, availability, results) to avoid “assumption lock-in” while also creating durable, well-scoped hook points for ISV contribution and clear boundaries between open stacks and vendor backends.
Theme 3, scheduling, allocation, and lifecycle operations, argues that time-to-solution is usually dominated by orchestration and queueing and not by quantum execution. The mismatch in time scales, from minutes to hours for scheduling versus milliseconds to seconds for runs, makes scheduler integration a non-optional architectural issue. It also notes that no single allocation model wins. System design should support multiple hybrid modes (co-scheduling, interleaving, malleability, time-multiplexing) and clearly separate device execution semantics from scheduler lifecycle control (acquire/release, accounting, policy).
Theme 4, reliability, portability, and evidence-driven convergence, treats drift, calibration cycles, and partial failures as first-class design constraints which needs standardized methods like nuanced health states (beyond up/down), telemetry, and logging as well as portability across multiple IRs and access modes without fragmenting semantics or provenance.
Aligned with the “HPC Connects” theme, this work connects stakeholders across facilities, vendors, and user communities to converge on pragmatic integration strategies. The poster provides a compact set of points and open questions realized in the EQS3 sessions. Past and future sessions intend to accelerate coordinated roadmap planning, reduce duplicated effort, and improve the credibility of HPCQC as a production-level capability over the years to come.
Contributors:
Format
on-demandon-site
