The Virtual MemComputing Machine, available as a SaaS platform, solves the largest and most complex industrial computations associated with optimization, big data analytics and machine learning. Proven to deliver the performance of quantum computing today, our clients are using MemComputing to solve problems that are impossible given classical or quantum methodologies.
Leveraging Our Disruptive Technology
Our proprietary Virtual MemComputing Machine (VMM)
is currently available in a Software as a Service platform, enabling the mathematical programming technology of our VMM to dramatically reduce costs, improve operational efficiencies, and increase profitability for our clients.
Start solving your hardest problems today, register to become a free Beta Tester!
Our Solvers
ILP
Integer Linear Programming
SAT
Satisfiability
Available for Demo
QUBO
Quadratic Unconstrained Binary Optimization
Available for Demo
MAX-SAT
Maximum Satisfiability
Available for Demo
Integer Linear Programming
The Virtual MemComputing Machine is an Integer Linear Programming Solver engine that is currently being used to solve industry’s most complex optimization problems. Clients have the choice of using a user interface to upload and run problems or an easy to integrate application programming interface (API) to integrate the service with their own systems.
ILP Optimization Modes
Dynamic Parameter Search
This mode allows users to extract near optimal sets of parameters for a given problem through its fully automated Markov Chain Monte Carlo algorithm, and returns the parameter distributions along with the best objective and variable assignments found.
Users can then apply these parameters to the Standalone Solver and the Dynamic Solution Search to improve the solution of the problem. The parameters found are typically considered “good” for problems similar in structure or nature, and can be used as starting parameters or boundaries in the Dynamic Solution Search mode.
Dynamic Solution Search
This mode represents the most basic usage of our Virtual MemComputing Machines. The purpose of this mode is to use the Markov Chain Monte Carlo algorithm to search for the optimal parameters to improve solution search.
If no optimal parameters ranges for the problem are known, the Dynamic Solution Search provides a tool to dynamically tune parameters while simultaneously improving solutions to the problem.
Standalone Solver
Users select this mode for a single MemCPU run. Whether it be to test a specific set of parameters or a known optimal set of parameters, this mode returns the number of unsatisfied constraints and objective versus time, as well as the best objective and variable assignments found.
AI Parameter Prediction
Coming soon! This mode will be able to predict the suitable parameters for a user’s problem and test them in a single run.
Our ILP Solver Supports
Easy to Use
File upload for manual testing
Binary & Integer Variables
(Continuous variables coming soon)
ILP Problems
As .MPS or CPLEX’s .LP files
CPU & GPU
Accelerated instances
Supported Modeling
and Programming Languages
AMPL
AIMMS
PYTHON
GAMS
*Will work with you on other automation/interface needs*
Professional Support
Here at MemComputing, your experience is our number one priority. Customers receive support from our optimization experts who are ready to help at a moment’s notice.
Contact us directly at [email protected] for additional assistance.