Even with the incredible power of today’s computers, there are still many complex problems these conventional systems can’t address. We have developed a revolutionary computation engine that has been verified by an independent laboratory to overcome these limits. We scale where others can’t – in fact, our performance scales linearly as the number of variables and constraints increase and optimal results are delivered orders of magnitude faster than other technologies.

Let us show you how MemComputing can help solve your hardest optimization problems, faster.


Founded in 2016 with the goal of making a computing system that functions more like the human brain, MemComputing is the result of years of research in unconventional computing by two PhD Physicists from UC San Diego, Max Di Ventra and Fabio L. Traversa. These researchers first introduced the mathematical concept of Universal Memcomputing Machines , and have shown that these machines can solve hard problems efficiently. They then introduced the concept of “Self-Organizing Logic Gates” to realize them in practice. The result is a class of digital (hence scalable) memcomputing machines that can be realized in both software and hardware using current technology.

MemComputing is a privately held company with offices in San Diego, California and is proud to be associated with EvoNexus and SDVG.


Machine Learning
Network Science & community detection
Scheduling, routing & logistics
Image / pattern recognition & anomaly detection
Graph theory applications
Trading strategies & portfolio optimization
Utilities & Energy
Software / hardware verification and validation
Protein modeling

MemComputing’s technology is ideally suited for computationally heavy optimization tasks in areas like machine learning, software / hardware design validation and scheduling / logistics to name a few. These types of problems can be extremely hard to solve, and can generate enormous value if optimal approximations can be readily computed.

MemComputing quickly delivers optimal results and pushes far beyond current approximation limits faced by other technologies when addressing the most difficult optimization problems.

We look forward to working with you to solve what previously may have been unsolvable.


MemComputing incorporates a fundamental aspect of how the human brain works—the novel notion of “collective states”—to speed computation. Rather than processing information sequentially, as current computers do, MemComputing uses “collective states” of interconnected memprocessors (elementary components of memcomputing technology) . This collective behavior, enhanced by time nonlocality (memory) and strong spatial correlations , enable MemComputing to quickly manipulate large data sets simultaneously.

MemComputing’s first computational engine accelerates and enables methods for solving combinatorial optimization, machine learning, and system validation problems, to name a few. With the ability to incorporate tens of millions of variables and scale in linear time , it represents a breakthrough that leapfrogs what was previously achievable and stands to usher in a new realm of possibilities.

Mathematical Programming Optimization Benchmark

Time to solve for a given number of variables

This chart shows the relative performance of our engine compared with the fastest available solvers for a known hard version of a MaxSAT problem. Memcomputing performance reflects interpreted Matlab code compared to the compiled code of the other two solvers. Memcomputing’s engine has been verified to scale linearly up to tens of millions of variables using similar problems.

MemComputing systems use a new computing feature called “self-organizing logic gates” (SOLGs) to rapidly search for solutions to an optimization problem – a proprietary and fundamentally different approach than traditional combinatorial methods. These SOLGs are then assembled into active dissipative “self-organizing logic circuits” whose equilibria are associated with either optimal minima of the problem at hand, or represent approximations to the optimum far beyond the inapproximability limit of current methods. This scalable solution can be implemented as both software and hardware using current technology.

MemComputing, Inc. has developed capabilities to natively solve binary optimization problems, by formulating them as “Boolean Maximum Satisfiability”, or MaxSAT, problems. Problems can also easily be structured as “quadratic unconstrained binary optimization”, or QUBO, problems. Both of these are common representations of optimization problems in industry.

While providing a significant increase in computational speed and accuracy, MemComputing, when implemented in hardware, also has the advantage of being able to operate at room temperature unlike the extreme operating conditions required of some alternative technologies. Additionally, the physical components used to create this technology are all readily available as existing consumer electronics and, therefore are immediately ready to scale.


We are currently in Alpha Test. Users can submit problems via our SaaS portal in Conjunctive Normal Form (CNF), which ultimately represents the normal form of Boolean propositions for the problem’s variables and constraints. The system sends these values along with other user-specified parameters to the Engine. In a fraction of the time required by other solvers, far better results are returned to the user.

Future releases will expand both the available problem formats and input specifications.


OCTOBER 23, 2017

Evidence of an exponential speed-up in the solution of hard optimization problems

Optimization problems pervade essentially every scientific discipline and industry. Many such problems require finding a solution that maximizes the number of constraints satisfied. Often, these problems are particularly difficult to solve because they belong to the NP-hard class, namely algorithms that always find a solution in polynomial time are not known. Over the past decades, research has focused on developing heuristic approaches that attempt to find an approximation to the solution. However, despite numerous research efforts, in many cases even approximations to the optimal solution are hard to find, as the computational time for further refining a candidate solution grows exponentially with input size. ...

AUGUST 16, 2017

Topological Field Theory and Computing with Instantons

It is well known that dynamical systems may be employed as computing machines. However, not all dynamical systems offer particular advantages compared to the standard paradigm of computation, in regard to efficiency and scalability. Recently, it was suggested that a new type of machines, named digital –hence scalable– memcomputing machines (DMMs), that employ non-linear dynamical systems with memory, can solve complex Boolean problems efficiently. This result was derived using functional analysis without, however, providing a clear understanding of which physical features make DMMs such an efficient computational tool. Here, we show, using recently proposed topological field theory of dynamical systems, …

MAY 7, 2017

Memcomputing Numerical Inversion With Self-Organizing Logic Gates

We propose to use digital memcomputing machines (DMMs), implemented with self-organizing logic gates (SOLGs), to solve the problem of numerical inversion. Starting from fixed-point scalar inversion, we describe the generalization to solving linear systems and matrix inversion. This method, when realized in hardware, will output the result in only one computational step. As an example, we perform simulations of the scalar case using a 5-bit logic circuit made of SOLGs, and show that the circuit successfully performs the inversion. Our method can be extended efficiently to any level of precision, since we prove that producing n-bit precision in the output …

FEBRUARY 8, 2017

Polynomial-time solution of prime factorization and NP-complete problems with digital memcomputing machines

We introduce a class of digital machines, we name Digital Memcomputing Machines, (DMMs) able to solve a wide range of problems including Non-deterministic Polynomial (NP) ones with polynomial resources (in time, space, and energy). An abstract DMM with this power must satisfy a set of compatible mathematical constraints underlying its practical realization. We prove this by making a connection with the dynamical systems theory. This leads us to a set of physical constraints for poly-resource resolvability. Once the mathematical requirements have been assessed, we propose a practical scheme to solve the above class of problems based on the novel concept …

FEBRUARY 13, 2015

Universal Memcomputing Machines

We introduce the notion of universal memcomputing machines (UMMs): a class of brain-inspired general-purpose computing machines based on systems with memory, whereby processing and storing of information occur on the same physical location. We analytically prove that the memory properties of UMMs endow them with universal computing power (they are Turing-complete), intrinsic parallelism, functional polymorphism, and information overhead, namely, their collective states can support exponential data compression directly in memory. We also demonstrate that a UMM has the same computational power as a nondeterministic Turing machine, namely, it can solve nondeterministic polynomial (NP)-complete problems in polynomial time. However, by virtue …


DECEMBER 06, 2017

Webinar: "The Science Behind MemComputing"

OCTOBER 03, 2017

"Interview with MemComputing CEO, John Beane"

SEPTEMBER 18, 2017

Memcomputing compute engine rivals Supercomputer Processing Speeds

Improve the Performance of Machine Learning Optimization Calculations by 4 Orders of Magnitude or More.

San Diego, CA, September 18, 2017– MemComputing Inc. is launching the first of many software-based coprocessors that solve complex problems that evade even the fastest supercomputers at TechCrunch DISRUPT. MemComputing’s first coprocessor, known as Falcon, is tuned specifically for machine learning and artificial intelligence. Falcon is free to use for a limited time as a cloud-based software as a service (SaaS) platform. Custom software development kits (SDK) are also available.

“We’re looking forward to our chance to showcase our technology at TechCrunch DISRUPT,” said John Beane, MemComputing Inc. CEO. “We will be showing other startups and companies how they can solve their most complex problems thousands of times faster than the current best processing systems.”
MemComputing’s technology is on the verge of disrupting machine learning, artificial intelligence and other industries where exceedingly long computation cycles are common. Come see MemComputing in the AI Startup Alley at TechCrunch DISRUPT today.

Mr. Beane added, “Our performance is really hard to fathom. Most people are very skeptical when they first learn of what we can do, especially when we tell them it’s done in software. Everyone expects that there is special hardware and probably a supercomputer behind it all. But that’s not the case. In fact, that is the main reason we made it so easy to access over the Internet with an initial period of free use. You don’t have to believe what we say. You can try it and see for yourself.”

JULY 6, 2015

New Brain-Like Computer May Solve World's Most Complex Math Problems

A new computer prototype called a "memcomputer" works by mimicking the human brain, and could one day perform notoriously complex tasks like breaking codes, scientists say.

These new, brain-inspired computing devices also could help neuroscientists better understand the workings of the human brain, researchers say.

In a conventional microchip, the processor, which executes computations, and the memory, which stores data, are separate components. This constant relaying of data between the processor and the memory consumes time and energy, thus limiting the performance of standard computers.

JULY 3, 2015

UCSD scientists build brain-inspired computer

Inspired by the human brain, UC San Diego scientists have constructed a new kind of computer that stores information and processes it in the same place. This prototype "memcomputer" solves a problem involving a large dataset more quickly than conventional computers, while using far less energy, the scientists say in a study...


Fabio Lorenzo Traversa

Chief Technology Officer

John A. Beane

Chief Executive Officer

Massimiliano Di Ventra



We are physicists, entrepreneurs, developers and inventors. We embrace challenging the status-quo to solve the previously unsolvable and to help our clients make better business decisions more quickly than ever before.

We are looking to expand our team with curious, talented individuals -- please tell us about yourself!