Notification from the conference:
We are pleased to inform you that your submission has been accepted for ORAL presentation and publication in the proceedings of SOICT 2025.
Paper Title: Toward Acceleration of Variational Quantum Classifier Simulation on GPUs
Author(s): Lan Anh Nguyen, Thai Anh Vo, Trang Nguyen, Son-Trung Doan, Toan-Duc Nguyen, Son-Hong Ngo, Eunsung Jung
Abstract:
The Variational Quantum Classifier (VQC) is one of the most extensively studied models in Quantum Machine Learning (QML). However, given the current limitations of quantum hardware, simulating QML algorithms on classical computing platforms such as CPUs, GPUs, and FPGAs has become an essential step for evaluating their performance and feasibility prior to deployment on real quantum devices. In this work, we present an accelerated VQC simulation framework, termed A-VQC, which exploits the parallel processing capabilities of classical hardware, particularly GPUs, to enable efficient quantum-classical simulation. Specifically, A-VQC incorporates two complementary acceleration strategies: (1) data-worker concurrency, which enhances data throughput by performing asynchronous and independent data-loading operations in parallel with VQC execution; and (2) stream-wise concurrency, which utilizes multiple GPU streams to train VQC mini-batches concurrently. A-VQC is implemented through a cross-platform integration of PennyLane and PyTorch. Experimental results show that A-VQC achieves approximately 10% faster training and 30% higher GPU utilization compared to conventional VQC simulation approaches.