Insights
Ideas worth exploring
Optimization is an unusually rich meeting place for mathematics and practical decisions. A small change in the way a problem is represented can alter what is computationally possible. A faster search can make an unfamiliar business choice visible in time to matter.
A quantum experiment can reveal as much about the quality of a classical baseline as it does about a new device.
Featured reading paths
The articles here pursue those ideas in depth. Choose a question or follow a carefully arranged series, then keep reading where the subject leads.
Business to Mathematics
Five connected essays about turning ambiguous priorities into variables, constraints and meaningful objectives.
Begin with What should the optimizer want?
Inside QUBO Search
Discover how interaction graphs, search landscapes, coefficients, solution quality and sensitivity shape difficult binary optimization.
Begin with The graph hiding inside your QUBO
Better Algorithms
Explore why strong solvers combine methods, adapt to instances and decide where to spend another second of computation.
Begin with The future of optimization is a team of algorithms
Plans Under Pressure
Consider the difference between a mathematically attractive schedule and a workable decision under changing conditions.
Begin with The Schedule Broke. Now What?
Quantum Compilation and Applications
Follow the journey from an abstract algorithm to a hardware experiment, including what its results can actually establish.
Begin with The Hidden Decisions in Quantum Circuit Compilation
Most recently published
- Insight
How good is an answer when nobody knows the optimum?
A QUBO solver can return an impressive number in seconds. Understanding what that number proves takes a little more mathematics.
Read the insight - Insight
Quantum-inspired optimization without a quantum computer
The mathematics of spins, energy and physical dynamics is helping invent new classical algorithms. What makes them useful is what they solve.
Read the insight - Insight
The graph hiding inside your QUBO
Two binary models can contain the same number of variables yet demand very different amounts of computation. Their interactions explain much of the difference.
Read the insight - Insight
The hidden landscape of a QUBO problem
Why a promising solution can be surrounded by worse ones, and what good algorithms do when the obvious moves run out.
Read the insight - Insight
The optimizer that chooses how to optimize
Modern solvers make decisions about their own search. Increasingly, those decisions can be learned.
Read the insight - Insight
Two equally good answers. Two different decisions
A QUBO can have many excellent solutions. The most valuable one may depend on what the objective does not capture.
Read the insight
Browse every Insight
One catalog, shared with Explore: search it, filter it and sort it there.