An EvoNexus Incubator Company
Moore’s law equates to CPU performance doubling every 18 months. However, Big Data are exploding exponentially. Analyzing these data to find the best solution from all possible solutions is classified as an optimization problem. Optimization problems are common in industries such as bioinformatics, autonomous vehicles, finance and transportation to name a few. Computing the results with data of this scale along with the associated constraints and conditions result in computational times that are growing beyond reason. So much so, that companies must turn to heuristics or algorithms that short cut the calculations and result in approximate solutions. In some cases, it is not possible to short cut the solution and/or an approximation would not be appropriate. Even with computational performance doubling every 18 months, classical computing methods just can’t keep up.
The MemComputing Solution
MemComputing, Inc. was formed around an entirely new computational paradigm invented by two PhD Physicists from UCSD. In classical computing, all calculations run iteratively in one direction. MemComputing is designed using computational memory that all work together cooperatively, much like the neurons in the brain. The result is a computation that is solved in a single cycle. Thus, problems are solved hundreds and thousands of times faster. In truth, MemComputing is exponentially faster than today’s best solutions. This seemingly impossible feat is real, it is available and has been verified by the San Diego Supercomputer Center using no more than a single thread from a Xeon processor.