NVIDIA cuOpt Managed Service

NVIDIA cuOpt Managed Service

• Overview
• Introduction
  
  • What is NP-Hard?
  • The Necessity for Heuristics
  • How cuOpt Solves the Problem
  • GPUs Unleash Massive Parallel Computing Capabilities
  • Approaches to Using cuOpt
• Quick Start
  
  • Step 1: NGC Account Setup
  • Step 2: Setup Python Environment
  • Step 3: Download Python Thin Client
  • Step 4: Obtain Credentials for cuOpt Managed Service
  • Step 5: Run Your First Sample Workload
  • Step 6: Sample Run Using the cuOpt CLI
  • Step 7: Sample Run Using API
• Data Requirements
  
  • Cost Matrix
  • WayPoint Graph
  • Fleet Features
    
    • Vehicle start and end locations
    • Vehicle Time Windows
    • Vehicle Breaks
    • Vehicle Priorities
    • Max Wait Time
    • Skip First – Drop Return Trips
    • Heterogeneous Fleet
    • Minimum Constraints per Vehicle
    • Maximum Constraints per Vehicle
  • Task Features
    
    • Pickup Deliveries
    • Order Priorities
• API
  
  • Troubleshooting
    
    • Errors and Causes
• Transition Guide for Change In Features
  
  • feasible/infeasible solution
  • solver_config
  • fleet_data
  • task_data
• Additional materials
  
  • Vehicle breaks
  • Feasibility/Infeasibility
  • Variance, Determinism and Solver time
  • Multi Trip VRP
• Key Concepts and Best Practices
  
  • Components of a Routing Optimization Problem
    
    • Environment
    • Tasks
    • Fleet
  • Introducing Additional Constraints
    
    • Task Time Windows
    • Mixed Fleet
    • Minimum Vehicles
    • Vehicle Priorities
    • Max Cost Per Vehicle
    • Prizes
    • Other Constraints
  • Avoiding Common Mistakes
    
    • Over Constrained Problems
    • Dimension Match
    • Time Value Consistency
• FAQ