Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

Quantum Circuits’ CUDA-Q integration links NVIDIA’s hybrid quantum-classical development platform with Quantum Circuits’ Aqumen software. When the companies announced the work on September 2, 2025, developers could prototype and test CUDA-Q applications in Quantum Circuits’ AquSim simulator; execution on the company’s Aqumen Seeker system and later quantum processing units (QPUs) was described as a future release. Quantum Circuits later said it had demonstrated Seeker with CUDA-Q, but its blog does not show a publication date.

What does the CUDA-Q integration enable?

The integration is intended to let developers build quantum applications with CUDA-Q and use Quantum Circuits’ Aqumen software and systems as part of a hybrid workflow. In its September 2, 2025 announcement, Quantum Circuits said the integration included core gates for universal quantum computing and real-time feedforward for conditional statements. These are features described by the company, not independent performance findings.

Quantum Circuits called it the first integration of CUDA-Q with a dual-rail programming environment using error detection. That “first” characterization is the company’s claim.

NVIDIA describes CUDA-Q as an open-source platform with a kernel-based programming model spanning CPU, GPU and QPU resources. Its current developer page lists Python and C++ programming models, GPU-accelerated simulation and QPU-agnostic support, and gives pip install cudaq as a getting-started command. NVIDIA also says CUDA-Q integrates with 75% of publicly available QPUs; the page does not display a publication year or methodology for that figure, so it should be read as NVIDIA’s undated claim rather than an independently verified statistic. NVIDIA CUDA-Q developer page

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

Can CUDA-Q run on Quantum Circuits hardware now?

The availability story has two distinct stages. The September 2025 announcement said developers could prototype and test CUDA-Q applications on AquSim, while execution on Aqumen Seeker and later QPU generations would come in a future release. It did not announce general Seeker execution as available at that time. Quantum Circuits’ September 2, 2025 announcement

In a later company blog, Quantum Circuits said it demonstrated its Seeker dual-rail-qubit system with CUDA-Q at NVIDIA GTC in Washington, D.C. The reviewed blog page has no visible publication date, and a demonstration does not by itself establish broad user access or general availability. Quantum Circuits’ Seeker demonstration blog

What is AquSim for?

AquSim is Quantum Circuits’ simulator for developing and testing applications before running them on hardware. The company describes it as error-aware, intended to model its dual-rail system and support algorithm exploration up to 25 qubits. That 25-qubit figure is a product description from Quantum Circuits; the blog does not provide an independent benchmark. Quantum Circuits’ Seeker demonstration blog

The intended progression is:

  1. Develop: Write quantum kernels and hybrid application logic with CUDA-Q.
  2. Prototype: Use AquSim to explore and test programs against Quantum Circuits’ modeled system.
  3. Execute: Run on Aqumen Seeker or a later QPU generation when the relevant hardware support and access are available.

The available descriptions do not specify access conditions, comparative simulator fidelity, hardware execution benchmarks, pricing or broad commercial availability. They therefore support understanding the workflow, but not a performance comparison between AquSim and a physical QPU.

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.

What does “dual-rail” mean in this announcement?

Quantum Circuits describes its dual-rail qubits as using a proprietary superconducting cavity architecture with built-in error detection. Error detection is a design claim from the company; the cited materials do not independently validate it or establish a head-to-head performance advantage over other qubit systems. The announcement’s focus on error detection also should not be conflated with full fault tolerance: the materials do not say this integration makes the system fault-tolerant.

Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Support on Ko-Fi

How does this fit with later CUDA-Q development?

NVIDIA announced CUDA-Q Logical on September 14, 2026, describing it as an orchestration layer for designing and testing fault-tolerant quantum applications. NVIDIA said Fermilab accelerated a fault-tolerant architecture workflow from five months to three weeks using CUDA-Q Logical, which the release characterized as a 7x speedup. This is a separate CUDA-Q development and is not a result of the Quantum Circuits integration. NVIDIA’s September 14, 2026 CUDA-Q Logical announcement

Quantum Circuits CEO Ray Smets said the partnership “accelerates the path to commercial-ready quantum computing.” NVIDIA’s Tim Costa, then senior director of Quantum and CUDA-X, said, “Hybrid quantum-classical systems are charting the path to useful quantum applications, and CUDA-Q is leading the charge.” These are the executives’ characterizations in the original announcement, not evidence that commercial readiness or useful applications have already been achieved.

Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.

Special offer. See more information about Outbyte and uninstall instructions. Please review EULA and Privacy policy.