Quantum Art, a developer of full-stack quantum computers based on trapped-ion qubits, announced achieving 10X compression in circuit depth and a 30% reduction in error rates by compiling circuits with its all-to-all connected multi-qubit gates on NVIDIA accelerated computing using the NVIDIA CUDA-Q platform. The improvements were verified in simulation on the CUDA-Q quantum-classical integration framework, underscoring the promise of combining Quantum Art's hardware-aware compilation with the NVIDIA accelerated computing ecosystem.
Quantum Art's fully programmable, all-to-all connected multi-qubit gates and advanced compiler serve as a critical resource for implementing circuits at smaller depth, enabling faster runtime and higher performance, thereby shortening the path to commercial applications at scale. The company's general-purpose compiler automatically optimizes input circuits and substitutes standard operations with efficient multi-qubit gates, consistently delivering order-of-magnitude compression and substantial performance gains.
“We designed our architecture to deliver real performance gains,” said Dr. Tal David, CEO of Quantum Art. “Programmable all-to-all multi-qubit gates are a critical advancement that supports our long-term goal of fault tolerant, commercially viable quantum computing.”
“Our compilation technique demonstrates how our multi-qubit gates and optimized compilers can compress quantum circuits by an order of magnitude while simultaneously improving performance by 30%,” said Dr. Amit Ben-Kish, CTO and co-founder of Quantum Art. “The general-purpose compiler optimizes very large quantum circuits with few multi-qubit gates. This compilation is verified by using the NVIDIA CUDA-Q platform to operate NVIDIA AI infrastructure.”
“By allowing researchers to draw on accelerated computing for their work, NVIDIA CUDA-Q is enabling next-generation breakthroughs in quantum computing,” said Sam Stanwyck, Group Product Manager for quantum computing at NVIDIA. “Quantum Art’s use of CUDA-Q to achieve circuit depth compression and error reduction is a clear example of how meaningful performance improvements are being realized by drawing on the latest advances in AI supercomputing.”
This breakthrough further validates and aligns with Quantum Art’s broader roadmap, which centers on scaling multi-qubit gates and reconfigurable multi-core architectures to deliver increasingly powerful quantum systems. The results were verified using the CUDA-Q integration announced earlier this year.
