AWS and JPMorganChase Study Smaller Inputs for Quantum Optimization
Their research combines classical preprocessing and problem decomposition with quantum workflows to reduce the burden placed on hardware.
D-Wave Two at NASA Advanced Supercomputing, used as broader optimization-hardware context. It was not the AWS–JPMorgan test platform. Image credit“D-Wave Two quantum computer inside the NASA Advanced Supercomputing Facility” by Oleg Alexandrov, CC BY-SA 4.0. · https://creativecommons.org/licenses/by-sa/4.0/
AWS and JPMorganChase are studying how to give quantum hardware a smaller problem to solve. Their collaboration report describes three studies using classical preprocessing and problem decomposition for quantum optimization. Reducing the workload before quantum execution can make limited hardware useful for larger original tasks.
In one portfolio-optimization workflow, the report describes about an 80% reduction in problem size and a threefold reduction in solution time on the tested instances.
The figures belong to that specific workflow and its comparisons. They neither establish a general quantum speedup nor remove the cost of classical preprocessing. A fuller assessment would include preprocessing time and quantum execution, with complete workflows measured against classical alternatives on representative instances.