D-Wave Simulation Quality Would Take 1 Million Years on Frontier Supercomputer
D-Wave responds to earlier claims of classical algorithm that can beat the quantum annealer
A simulation performed by D-Wave’s quantum processing unit would take nearly a million years to replicate on the Frontier supercomputer, according to results published in Science and now reaffirmed by D-Wave Quantum Inc. The company is responding to claims that recent classical simulation work has invalidated its demonstration of beyond-classical computation in quantum simulation.
Researchers at the Flatiron Institute have made progress with their BP-TNS algorithm by incorporating large diamond-lattice data into their simulations, but this work falls short of fully replicating D-Wave’s original experiment, failing to compute the same observables, geometries, sizes, or couplings. “These are important cases from our Science paper, and they remain beyond the reach of current classical methods,” said Dr Trevor Lanting, chief development officer at D-Wave, adding that claims to the contrary are inaccurate without reproducing the full scope of the original demonstration.
D-Wave Demonstrates Beyond-Classical Computation in Quantum Simulation
Recent assertions that classical simulations have invalidated D-Wave’s quantum supremacy claims are inaccurate; the company maintains its 2024 Science publication demonstrates computational advantages in specific quantum simulations. D-Wave asserts that matching the quality of its simulations to conventional methods would require approximately 1 million years of processing time on the Frontier supercomputer, along with impractical energy and memory requirements. This claim centres on the quantum simulation of nonequilibrium magnetic spin dynamics, a field where D-Wave’s annealing quantum computers have reportedly surpassed classical capabilities. The Flatiron Institute’s work, utilising a BP-TNS algorithm and published in Science, has incorporated large diamond-lattice data into its simulations, but D-Wave emphasises that this doesn’t replicate the full scope of its original experiment. D-Wave highlights the limitations of the competing research.
Further analysis, detailed in an arXiv paper, revealed that BP-TNS struggles with strongly coupled three-dimensional spin glasses on cubic and diamond lattices, critical elements of D-Wave’s initial demonstration. Dr Trevor Lanting, chief development officer at D-Wave, stated, “These are important cases from our Science paper, and they remain beyond the reach of current classical methods.” D-Wave welcomes continued advancements in classical simulation, but stresses the importance of accurately characterising scientific results and of avoiding claims to have overturned established quantum supremacy demonstrations.
BP-TNS Algorithm Limitations on 3D Spin Glass Problems
The pursuit of quantum supremacy continues to drive innovation in both quantum and classical computing, with recent attention focused on simulating complex magnetic spin dynamics. While D-Wave Quantum Inc. maintains its 2024 Science publication demonstrates beyond-classical computation, researchers at the Flatiron Institute have been actively developing classical algorithms to challenge this assertion. Their work centres on the BP-TNS algorithm, a tensor-network method for simulating these systems, but its limitations in applicability are becoming increasingly apparent when compared with the scope of D-Wave’s initial experiments. Specifically, the Flatiron Institute’s advancements, detailed in a recent Science publication adding “large diamond lattice data,” do not fully replicate the conditions of D-Wave’s simulations. D-Wave researchers demonstrated their results using square, cubic, diamond, and biclique topologies, showing quantum processing units produced samples consistent with quantum theory at scales where classical computation becomes impractical.
Matching the quality of D-Wave’s simulation on the largest problems, the company claims, would require nearly a million years on the Frontier supercomputer. “The BP-TNS algorithm is effective in some regimes and ineffective in others,” said Dr Trevor Lanting, chief development officer at D-Wave. “Our analysis showed that it fails for strongly coupled three-dimensional spin glasses on cubic and diamond lattices, and that loop-corrected BP-TNS is ineffective for higher-dimensional biclique problems. These are important cases from our Science paper, and they remain beyond the reach of current classical methods.”
Frontier Supercomputer Inability to Match D-Wave Results
Researchers continue to assess the claims of quantum advantage put forth by D-Wave Quantum, with recent classical simulation efforts failing to fully replicate the company’s published results. While the Flatiron Institute has made advancements in classical simulation techniques, specifically with its BP-TNS algorithm and the addition of large diamond lattice data, a complete match to D-Wave’s experiments remains elusive.
The Flatiron team’s work, detailed in a recent Science publication, focuses on extracting a Kibble-Zurek exponent in a three-dimensional system, but doesn’t encompass the full scope of D-Wave’s initial demonstration. Dr Trevor Lanting, chief development officer at D-Wave, stated, “These are important cases from our Science paper, and they remain beyond the reach of current classical methods.”
D-Wave’s Dual-Platform Quantum Computing Capabilities
D-Wave Quantum continues to refine its position as a unique player in the quantum computing field, offering both annealing and gate-model systems, a dual-platform approach that sets it apart. While much attention focuses on the race to build increasingly powerful quantum computers, D-Wave asserts its existing systems (quantum annealers) already demonstrate capabilities beyond the reach of even the most powerful classical supercomputers for specific tasks.
Your newsletters always make me stretch🥀