Qoro Quantum
CESGA and Qoro Quantum use high-performance computing to model distributed quantum circuits. Scalable, distributed quantum circuit simulations across ten HPC nodes were demonstrated in a pilot project conducted by Qoro Quantum and CESGA that combined Qoro’s orchestration software with CESGA’s CUNQA emulator. Thousands of quantum circuits were produced and scheduled for simulation on CESGA’s infrastructure using Qoro’s Divi software to evaluate distributed VQE and QAOA implementations.
Workloads for VQE and QAOA completed in less than a second, proving that high-throughput quantum algorithm simulations can be carried out with little code and effective use of resources.
The pilot demonstrated the potential of distributed emulators like CUNQA to get ready for upcoming large-scale quantum computing deployments by confirming the viability of hybrid quantum-classical operations in HPC systems.
The Galician Supercomputing Centre (CESGA) and Qoro Quantum have finished a pilot study that shows how scalable, distributed quantum circuit simulations may be supported by high-performance computing systems. The two-week collaboration centred on implementing Qoro’s middleware orchestration platform to execute distributed versions of two essential quantum algorithms the variational quantum eigensolver and the quantum approximate optimisation algorithm across CESGA’s QMIO infrastructure, according to a Qoro Quantum release.
HPC Systems and Quantum Workload Integration
Divi, a quantum application layer in Qoro’s software package, is intended to automate the orchestration and parallelization of hybrid quantum-classical algorithms. Divi prepared and performed quantum workloads on 10 HPC nodes utilising CESGA’s distributed QPU simulation framework CUNQA for the pilot.
According to the release, CUNQA from CESGA offers a modular testbed that simulates distributed QPU environments with adjustable topologies and noise models. This made it possible for Qoro’s platform to replicate the demands of upcoming hybrid quantum-HPC systems by simulating actual quantum workloads in a multi-node arrangement.
Everything interacted flawlessly, the cooperation went very smoothly, and the end-to-end functionality performed as planned.
Examining QAOA and VQE in a Distributed HPC Setting
One of the main issues in quantum chemistry is estimating the ground-state energy of quantum systems using the variational hybrid technique known as VQE. In this pilot, Qoro and CESGA used two ansätze Hartree-Fock and Unitary Coupled Cluster Singles and Doubles to model a hydrogen molecule. Divi produced 6,000 VQE circuits in accordance with the bond lengths, which were changed between 20 values.
The CUNQA emulator explored the ansatz parameter space using Monte Carlo optimisation on 10 computing nodes. Qoro claims that it only took 0.51 seconds to replicate the full demand. With only 15 lines of Divi code, the platform can enable high-throughput experimentation, as evidenced by the data that were automatically collected and returned for examination.
Additionally, the researchers tested QAOA, a quantum-classical method used to solve Max-Cut and other combinatorial optimisation issues. This problem, which is related to data clustering, circuit design, and logistics, entails splitting a graph to maximise the number of edges between two subgroups.
A 150-node graph was divided into 15 clusters for the simulation, and parameterised circuits were created using Monte Carlo techniques using Qoro’s Divi program.Two test scenarios were run: 21,375 circuits 15.44 seconds and 2.13 seconds for 2,850 circuits. Sample size increased the cut size ratio, which compares quantum to classical results, from 0.51 to 0.65. Once more, the CUNQA emulator was used to run all circuits in parallel using CESGA’s infrastructure.
Results, Infrastructure, and Prospects
A number of concrete results from the pilot study show advancements in scalable hybrid quantum computing. Qoro’s orchestration platform and CESGA’s distributed quantum emulator allowed flawless communication between the simulated QPU infrastructure and application layer, according to the Qoro Quantum release. The partnership also showed how to use Qoro’s Divi software to automatically create and plan massive quantum workloads, which would simplify the execution of intricate quantum programs.
Furthermore, the experiment demonstrated that distributed execution can greatly boost speed without requiring a lot of manual setup by validating the viability of executing hybrid quantum-classical algorithms across numerous HPC nodes. Lastly, the pilot revealed important technological factors that should be taken into account when scaling quantum workloads in high-performance computing settings. These findings will guide the creation of distributed quantum systems in the future.
The statement states that by simulating distributed quantum architectures, the experiment demonstrates how existing HPC infrastructure can handle future quantum workloads. In order to facilitate the wider implementation of quantum computing in expansive classical settings, Qoro Quantum and CESGA intend to keep improving this strategy.
With assistance from the European Union and the Spanish Ministry for Digital Transformation, CUNQA is being created as part of the Quantum Spain initiative. As part of the EU’s COVID-19 response effort, ERDF_REACT EU provided funding for the QMIO infrastructure utilized in this project.
2 thoughts on “Qoro Quantum And CESGA For Distributed Quantum Simulation”