Embedded Files
Moonlight Technologies Announces "Quantum Optimization Framework" Feature for EvoSpikeNet
Yokohama, August 31, 2026 – Moonlight Technologies Inc. (CEO: Masahiro Aoki; Headquarters: Yokohama) has announced an optimization framework for its distributed neuromorphic framework, "EvoSpikeNet," designed to seamlessly utilize both quantum and classical computers. Featuring an architecture that explicitly decouples mathematical models, data structures, and execution parameters, developers can define a problem once and build applications without having to worry about the underlying execution device (QPU, simulator, or CPU/GPU).
Background Combinatorial optimization problems (such as route planning, resource allocation, and scheduling) can be solved using various methods, including quantum annealing, QAOA (Quantum Approximate Optimization Algorithm), and classical simulated annealing. However, each method requires a different input format and API, forcing developers to write custom code for each specific target hardware (IBM Quantum, D-Wave, OpenJij, Fixstars Amplify, classical simulators, etc.). EvoSpikeNet solves this issue by separating the architecture into three layers: problem definition, data structure (QUBO conversion), and execution backends.
Feature Overview
Device-Agnostic Problem Definition (OptimizationProblem): Enables hardware-independent modeling of key elements such as candidates, conflict_pairs, required_groups, and dynamic_constraints. It supports multiple domains, including road networks, low-/high-altitude flight paths, and general assignment/scheduling tasks.
QUBO Conversion Layer (QuboProblemBuilder): Automatically converts problem definitions into QUBO (Quadratic Unconstrained Binary Optimization) matrices compatible with quantum annealing and quantum optimization backends. Hard and soft constraints can be flexibly incorporated as penalty terms.
Pluggable Execution Backends: Supports backends like classical simulated annealing, QAOA simulators, Qiskit (IBM Quantum), D-Wave, OpenJij, and Fixstars Amplify. Backends can be swapped at runtime simply by specifying their name. If quantum hardware is unavailable, the system automatically falls back to classical execution, ensuring continuous application uptime.
Unified Solution Contract: Regardless of the backend used, solutions are returned in a standardized format containing assignment, score, and meta (execution details), unifying downstream application logic.
Benefits for Developers This framework allows developers to focus on domain-specific logic without being bogged down by the details of optimization algorithms or underlying hardware. It enables a smooth, phased migration across the same codebase—allowing teams to prototype on classical simulators and seamlessly switch to quantum hardware in production.
Python
from evospikenet.optimization import OptimizationProblem, QuboProblemBuilder
from evospikenet.optimizer import get_optimizer
problem = OptimizationProblem(
name="resource-allocation",
domain="generic",
candidates=[{"id": "task-a", "cost": 4.0}, {"id": "task-b", "cost": 7.0}],
conflict_pairs=[("task-a", "task-b")],
)
qubo_problem = QuboProblemBuilder().build(problem)
solver_input = qubo_problem.as_problem_dict()
# Execute the same problem definition on different backends simply by switching devices
result = get_optimizer("qiskit").solve(solver_input)
# result = get_optimizer("classical").solve(solver_input)
For detailed specifications, please refer to the QAOA documentation site. Note that this standard represents the Core-side design specification.
Executive Statement Masahiro Aoki, CEO of Moonlight Technologies Inc., stated:
"While practical quantum computing is still evolving, it is critical to provide an environment where application developers can start building today with future hardware advancements in mind. EvoSpikeNet's optimization framework serves as that bridge."
Company Overview
Company Name: Moonlight Technologies Inc.
CEO: Masahiro Aoki
Established: 2025
Business Activities: AI framework development, distributed brain simulation, and neuromorphic computing
Website: https://www.moonlight-tech.biz
Contact Information
Email: info@moonlight-tech.biz
For further details regarding this press release, please reach out using the contact information above.
Page updated
Report abuse