Oracle Corp. in recent years has sharply ramped up its investment in Nvidia GPUs as it beefed up its Oracle Cloud Infrastructure to handle the growing demand of enterprise AI. This week, those two companies and another partner called QMware announced they are collaborating with a new direction in mind–hybrid quantum-classical computing.
Nvidia and Oracle are teaming with Swiss firm QMware, which describes itself as a “hybrid quantum computing company,” to use the Oracle Cloud Infrastructure (OCI) powered by Nvidia A100 Tensor Core GPU clusters, along with Nvidia’s CUDA Quantum open-source hybrid quantum computing platform, to test and develop a hybrid quantum computing service for enterprise customers. An initial version of the service is scheduled to be demonstrated at Oracle CloudWorld London on March 14.
Nvidia and Oracle are longtime partners, with the latter buying plenty of GPUs and the former running its DGX Cloud AI training-as-a-service platform on the OCI. Oracle said back in late 2022 that it was in the midst of adding “tens of thousands” of Nvidia GPUs to the OCI as it tackled AI demand.
As Nvidia also moved into the realm of quantum computing over the last couple years, QMware was among its earliest partners, as the pair explored the potential of GPU-based quantum simulation and more. Oracle, for its part, has not commented much on the nascency of quantum computing or the recent acceleration toward hybrid quantum-classical computing, which combines quantum techniques and software with classical hardware like GPUs. That may not be surprising for a company that also was somewhat cautious about embracing cloud computing and AI as excitement grew around those technologies.
Now, the new service enabled with QMware’s help could allow Oracle customers to dip their toes into the waters of quantum computing by first helping them explore their own industrial use cases for hybrid quantum-classical computing. The partners said the accelerated computing capabilities provided by Nvidia A100 GPUs, coupled with OCI Compute bare-metal instances and RDMA cluster networking, will provide QMware with a diverse array of options for testing and developing commercially-valuable quantum computing applications that can be used in
fields such as AI, quantum machine learning, and quantum-enhanced optimization. Nvidia’s CUDA Quantum helps to enable quantum simulations on classical hardware, along with the ability to program hybrid CPU, GPU, and quantum processing unit (QPU) systems.
Martin Peck, Vice President of Technology Software Engineering at Oracle, said, “Hybrid quantum computing has the potential to reshape how businesses operate, gain insights from data, and innovate new products and services. QMware is a front-runner in this exciting field, and we are pleased to bring the power and flexibility of Oracle Cloud Infrastructure to support them in building hybrid quantum services for enterprises.”
QMware CEO and co-founder Markus Pflitsch added, “We are entering an era where quantum computing is transitioning from experimental to practical. Our collaboration with Oracle symbolizes a huge step in this journey. By combining QMware’s expertise in quantum technology with Oracle’s robust cloud infrastructure, we are scaling up quantum capabilities and simplifying its accessibility for businesses across all sectors.”
QMware CTO and co-founder George Gesek also noted, “The future of quantum computing lies in its integration into the fabric of daily business operations, and through this collaboration, we are helping to make this future a reality.”
Tim Costa, Director of Quantum Computing and HPC at Nvidia, concluded, “High-performance quantum simulation is crucial for researchers to tackle the toughest challenges in quantum computing. Through collaborations with innovators such as Oracle and QMware, Nvidia helps enable the world’s researchers to achieve breakthroughs toward useful quantum-integrated computing.”