How Big of a Quantum Computer is Required to Perform a Large-Scale CFD Simulation?
Abhishek Chopra
Founder & CEO of BQP | CFD & CAE Simulations | Quantum Computing | HPC |
Researchers estimate that a large-scale jet engine simulation which took 19.2M HPC cores can be conducted using a quantum computer with 30 logical qubits
I'm thrilled to share the BIG NEWS - BQP's seminal research paper on #QuantumCFD (Q-CFD) is out! This work puts forth a concrete milestone required in the maturity of #quantumcomputers to solve a critical and practical #engineering simulation problem.
Overview
In this research, we demonstrated the scalability and accuracy of BQPhy's Hybrid #Quantum-Classical Finite Method (#HQCFM) solver on a Quantum simulator to solve a Computational Fluid Dynamics (#CFD) simulation. HQCFM is integrated with a variational quantum algorithm and solves the 2D Burgers equation - a non-linear and time-dependent partial differential equation (#PDE).
Scalability and Accuracy study
The most significant part of this paper was the scalability study which was performed from 4-#qubits all the way to 11-qubits. Then the resource estimation was performed, using the results of the scalability study (qubits, gates, & circuit size), for a large-scale jet engine simulation. The classical simulation was performed on 19.2M HPC cores by the Gordon Bell award nominee last year. Our results showcased that this same simulation can be performed on a 30-logical qubit quantum computer and would require ~4000 quantum gates!!!
The solver showcased an accuracy range of 10^(-6) to 10 ^(-11). The results are one of the highest accuracy shown to date for quantum-based approaches for CFD.
How far are 30 logical qubits in reality?
This past week at Quantum World Congress 2024, Quantinuum unveiled its roadmap to achieving universal, fault-tolerant quantum computing by 2030 (shown below). They showcased that their Helios system with ~50 logical qubits can be realized as soon as 2025. That means in about a year, BQP's Q-CFD solver could solve large-scale CFD problems! Personally, this is overwhelming and exciting at the same time, as it was previously perceived that such systems are 3-5 years away.
领英推荐
Where are we based on NASA's CFD Vision 2030 Roadmap?
NASA's CFD Vision 2030 study stated that CFD on Qunatum systems will demonstrate a representative model problem by 2025!! Is it a coincidence or well thought out plan?
What is the BQP team gearing up for next?
Huge shoutout to Ferdin Sagai Don Bosco and Dhamotharan S for their excellent work and collaboration in putting together this significant work at the intersection of CFD and Quantum computing. Rut Lineswala and I, as CFD practitioners and visionaries, are super excited to continue to striving for our goal to simulate full aircraft with high fidelity (which has been stated by experts to not be possible until 2080) in the next few years! Thanks to Jash Minocha and Aditya Singh for their support.
What are people saying about this study?
“BQP’s results signal the introduction of drastically higher computing power to flow field analysis and simulation. This capability can unlock new methods in aerospace development, enabling higher confidence during design and more proactive maintenance during the aircraft life cycle,” said Dan Hart , Senior Aerospace Executive and Member of the National Academy of Engineering.
Founder - QRiver
6 个月Abhishek Chopra MIMIQ has processed 272 qubits exactly (16 logical qubits) for complex error correction codes. We can go all the way to the 30 log qubits mentioned in your post- how about giving it a try ? ????
<??Quantum + Quant??> @LTIMindtree
6 个月Great work
Founder & CEO of BQP | CFD & CAE Simulations | Quantum Computing | HPC |
6 个月Link to original post:
We are closer than some think!