NCSA Events Calendar

NCSA, in collaboration with IQUIST, invites you to the monthly Quantum User Group gatherings for the campus community.

The Quantum User Group (QUG) is being formed to bring together several efforts that have been underway and to raise awareness of activities of which members of the campus community may wish to take advantage. The QUG creates a forum for interaction and information sharing in order to facilitate growing the local community of researchers with interest in quantum computing and raise awareness about emerging quantum computing capabilities and for those exploring the use of hybrid classical/quantum computing to share their experiences--both successes and challenges--with others. 

We encourage researchers with challenging research problems that can potentially benefit from quantum computing along with those who are curious about this to help determine the potential for benefiting your research to participate.  Graduate students and postdocs affiliated with research in classical as well as quantum computing and all campus community members interested in quantum computing are also encouraged to attend.

Agenda:

Welcome

Lightning Talk - 15 min:

Amul Prakash

Title: Exploring Quantum Sound and Music

Abstract: This project explores how quantum computing can reshape the capabilities of modern

musical composition software. By encoding musical elements such as pitch and harmony into

quantum states and circuits, our project investigates the potential for quantum phenomena,

including superposition and entanglement, to generate and reshape classical music. Specifically,

research has uncovered new methods for correlated music variation by entangling pitch and

harmony qubits with ancilla qubits using controlled quantum gates.

Featured Talk - 30 min: 

Ayush Bansal

Title: Quantum-Classical Co-Design for Fault-Tolerant QCSC: A Memory-Centric Framework for Quantum Error Mitigation

Abstract: Outputs from current Quantum Processing Units (QPUs) are highly noisy, making Quantum Error Mitigation (QEM) and Correction (QEC) essential. However, QEM processes scale exponentially and are incredibly memory-hungry. As QEM techniques like Probabilistic Error Cancellation (PEC) and Zero-Noise Extrapolation (ZNE) scale, powerful GPU accelerators are routinely starved of data. Compounding this bottleneck, essential linear algebra routines like sparse eigensolvers suffer from critically low arithmetic intensity and cache-polluting strided memory accesses, trapping current hybrid systems behind a classical "memory wall".

To address this lack of compute-to-memory scaling optimization, we introduce a novel memory-centric framework designed to execute QEM post-processing tasks efficiently within classical compute constraints. This framework fundamentally shifts the hybrid computing paradigm from a compute-centric offloading strategy – where tasks are blindly dispatched to GPUs regardless of their arithmetic intensity – to a highly intelligent, data-aware task orchestration model.

We will detail the implementation of a three-tier, lightweight telemetry stack capable of achieving precise "dominant cost" attribution. By feeding this telemetry into a dynamic Space-Time-Fidelity cost model, we can autonomously route workloads, adjust error margins, and guarantee determinism across complex, multi-tenant QEM/QEC quantum-centric supercomputing pipelines.

Also, a quick reminder to sign up for Slack Channel in order to stay up to date on the QUG’s activities.

We look forward to seeing you!