Computer chipmaker Nvidia Corp. is beefing up its efforts to advance quantum computing with the debut of a new service aimed at researchers and developers.

According to the company, Nvidia Quantum Cloud provides advanced quantum computing technologies that will push the boundaries of research in areas such as chemistry, biology and materials science.

Nvidia Quantum Cloud was announced at the chipmaker’s annual GTC conference today. Available in early access now, it’s based on the existing Nvidia CUDA-Q Quantum computing platform, which is said to be used by almost three-quarters of organizations that have deployed so-called “quantum processing units” or QPUs.

Nvidia Quantum Cloud is being made available as a microservice, making it simple for customers to build and test new quantum applications and algorithms deployed on cloud infrastructure.

According to Nvidia, Quantum Cloud provides access to cuQuantum software that simulates QPUs on graphics processing unit-based hardware. Existing QPUs are not yet powerful or stable enough to test quantum algorithms and applications, so a simulator provides the best opportunity for developers to explore how this new technology will impact scientific exploration.

The company presented a number of example use cases for Quantum Cloud. For instance, it partnered with the University of Toronto on an initiative known as the Generative Quantum Eigensolver, which leverages large language models to help quantum computers find the ground-state energy of molecules more rapidly, aiding in molecular research.

In a second initiative, Nvidia is teaming up with the quantum computing software startup Classiq Technologies Inc. The integration makes it simpler for researchers to quickly create large, sophisticated quantum algorithms and analyze and execute quantum circuits. In addition, Nvidia has partnered with QC Ware Corp. on a project called Promethium, which is a molecular simulator that helps to tackle highly complex problems in chemistry.

Nvidia said the Quantum Cloud service will be offered on Amazon Web Services, Google Cloud, Oracle Cloud Infrastructure and Microsoft Azure, as well as the cloud platforms of quantum computing companies such as IonQ Inc., IQM Quantum Computers Inc. and Quantinuum Inc.

Tim Costa, Nvidia’s director of high-performance computing and quantum computing, said quantum represents the next frontier of computing, but making it a reality will require collaboration among the world’s most brilliant minds. “Nvidia Quantum Cloud breaks down the barriers to explore this transformative technology and lets every scientist in the world harness its power,” he said.

Advancing quantum research

Nvidia shared how a number of organizations have been using its quantum computing infrastructure in areas such as financial fraud detection and the creation of new quantum educational courses.

For instance, HSBC Holdings plc, one of the world’s largest banks, used Nvidia’s quantum platforms to develop a new, quantum-based machine learning algorithm that’s able to detect fraudulent digital transactions.

The bank’s researchers explained that they used the CUDA-Q quantum computing platform to simulate 165 qubits on Nvidia’s GPUs, far more than anyone has done before. It said it used machine learning techniques implemented with CUDA Quantum and cuTensorNet software to overcome the challenge of simulating quantum circuits at large scale.

The chipmaker also revealed it’s working with almost two dozen universities on a program that’s designed to prepare a new generation of computer scientists with quantum skills. “Bridging the divide between traditional computers and quantum systems is essential to the future of computing, said Carnegie Mellon University Vice President of Research Theresa Mayer. “Nvidia is partnering with institutions of higher education, Carnegie Mellon included, to help students and researchers navigate and excel in this emerging hybrid environment.”

Hybrid quantum/classical infrastructure

Nvidia’s quantum infrastructure is also aiding in the development of new supercomputer platforms that will be able to simulate future quantum systems more accurately.

The company announced a trio of new systems for quantum simulations, with the largest being ABCI-Q at Japan’s National Institute of Advanced Industrial Science and Technology (pictured below). When it’s completed, ABCI-Q will power high-fidelity quantum simulations for research in various fields. It’s said to be integrated with Nvidia CUDA-Q, which provides powerful simulation tools and capabilities for programming hybrid quantum-classical systems. It will integrate a whopping 2,000+ Nvidia H100 Tensor Core GPUs across more than 500 nodes, connected using Nvidia’s advanced Quantum-2 InfiniBand networking technology.

A second system based on Nvidia’s DGX SuperPOD server is being deployed by the Novo Nordisk Foundation in Denmark, where it will aid in that country’s national plan to advance quantum computing.

The third new system coming online soon is the previously announced Pawsey Supercomputing Research Center in Australia, which will combine Nvidia CUDA Quantum software with Grace Hopper Superchips.

Post-quantum cryptography

Elsewhere, Nvidia revealed how it’s striving to make progress in the area of quantum security, unveiling a new cuPQC library that leverages the parallelism of its new GPUs to create more powerful algorithms that can defend against quantum computing attacks.

For years, researchers have raised concerns that once a viable quantum computer comes available, it will be able to crack the standard public key technologies used to secure communications in a matter of seconds. As such, there has been a lot of focus on so-called quantum security.

Nvidia said that the U.S. National Institute of Standards and Technology is planning to introduce the world’s first post-quantum cryptography algorithms later this year. The standards were developed using Nvidia’s platforms.

Douglas Stebila, co-founder of the Open Quantum Safe project, a group spearheading work in the nascent quantum security field, said his organization recently used the Nvidia cuPQC library to accelerate an algorithm called Kyber by up to 500 times. It did this by running the algorithm on an Nvidia H100 Tensor Core GPU, rather than a standard central processing unit.

Nvidia said the performance will be accelerated even more when its new Blackwell architecture GPUs become available. “Securing data against quantum threats is a critically important problem, and we’re excited to work with Nvidia to optimize post-quantum cryptography,” Stebila said.

The Open Quantum Safe project is an initiative launched by the newly formed Post-Quantum Cryptography Alliance, which is hosted by the Linux Foundation and seeks to develop various security-focused, post-quantum applications and libraries. Cloud providers including Amazon Web Services Inc., Google Cloud and Microsoft Corp. are also contributing to post-quantum research with Nvidia, together with startups such as QuSecure Inc., EvolutionQ Inc. and PQShield Ltd.

Images: Nvidia