Zapata Computing anounces multi-GPU cluster integration with NVIDIA cuQuantum for accelerating enterprise quantum use cases

Integration of Orquestra® platform for building quantum-ready applications and NVIDIA cuQuantum quantum simulation SDK supercharges quantum and AI workflows

Zapata Computing, a leading enterprise software company for quantum solutions, today announced that a new multi-GPU cluster integration is now available to customers. The NVIDIA cuQuantum SDK of optimized libraries and tools for accelerating quantum computing workflows is now integrated with Zapata’s Orquestra platform for building quantum-ready applications.

The hardware integration uses autoscaling to provision and mount GPUs on demand, avoiding unnecessary costs by provisioning only when necessary. The cuQuantum software integration allows users to simulate quantum circuit runs on GPU-accelerated systems, providing more than a 100X improvement over conventional simulators and hardware, and allowing for scaling of circuit sizes, previously limited to 20 qubits, up to 30 qubits; equivalent to a system 1,000X larger than 20 qubits.

Work continues to incorporate this functionality into the next generation of the Orquestra SDK to improve the user experience and accelerate the R&D cycle.

This integration is part of Zapata’s larger collaboration with NVIDIA, which includes the use of QODA, the quantum-classical platform recently announced by NVIDIA. QODA is a hybrid toolchain that will provide a unified quantum-classical environment to help accelerate Zapata’s ongoing research and development of quantum machine learning and optimization algorithms, targeting enterprise use cases.

“We’re thrilled to be working with NVIDIA on this integration and the QODA hybrid toolchain initiative,” commented Gregg Carman, Chief Revenue Officer at Zapata. “NVIDIA-powered HPC and other compute-intensive environments and our Orquestra computational workflow platform form a powerful combination for large organizations that can’t afford to wait for the day when fully fault-tolerant quantum computers are mainstream. And they don’t have to, because the tooling to start gaining a competitive advantage is already here, as personified by the cuQuantum and Orquestra integration.”

“High-performance computing workflows today, especially in enterprise and government, simulate quantum computing jobs at scale and require performance that greatly exceeds what existing quantum devices are capable of,” said Timothy Costa, Director, HPC and Quantum Computing Product at NVIDIA. “While the calculus of when to use HPC and quantum will change as quantum computers become more powerful, classical computing will always have an important role, and this hybrid computing model will need a unified programming environment that’s efficient and easy to use — which is why an integration like the one we’ve built with Zapata will create real value now and in the future.”

Related News