Core Concepts
D-Wave's new Nonlinear-Program Hybrid Solver (NL-Hybrid) excels in solving optimization problems with complex variable types like permutations and subsets, outperforming existing methods on the Traveling Salesman Problem and Knapsack Problem, but showing less efficiency on problems best suited for binary variable encoding like the Maximum Cut Problem.
Stats
NL-Hybrid achieves near-optimal solutions for TSP instances with up to 439 nodes.
For MCP, BQM-Hybrid and CQM-Hybrid significantly outperform NL-Hybrid.