Skip to content

The Neuromorphic Computing Platform based on a new flexible blockchain protocol

Notifications You must be signed in to change notification settings

realsetvin/dynexcoin

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

23 Commits
 
 

Repository files navigation

Decentralized Computing Dynex [DNX]

World’s first neuromorphic supercomputing blockchain based on the DynexSolve chip algorithm, a Proof-of-Useful-Work (PoUW) approach to solving real-world problems.

DynexSolve: Proof-of-Useful-Work (PoUW)

Our proprietary Dynex Chip design is built based on ideal memristors. Memristors are two-terminal resistive devices with memory. In general, their nonlinear dynamic behaviour is mathematically modeled by means of a differential algebraic equation (DAE) set, in which an ordinary differential equation (ODE) governs the time evolution of the memory state, while an algebraic relation captures the state- and input-dependent Ohm law. The memristor, an acronym for memory resistor, was theoretically introduced in 1971 by L.O. Chua. Introduced in the 1971 pioneering paper, presently referred to as ideal memristor, is the fourth fundamental two-terminal circuit element, the other three being the resistor, the capacitor, and the inductor. Since then, the interest on memristors and their applications has been growing exponentially, with both academia and industry deploying a huge amount of funds and personnel to fabricate, model, and explore the full potential of these devices in electronics applications. The DynexSolve PoUW algorithm utilises the unprecedented performance of such memristors and performs ODE integration (simulations) of our Dynex Chips. By utilising the massive parallelism of all participating Graphic Processing Units (GPUs), we can achieve close to realtime performance of the original chip design. This allows computations of constraint satisfaction problems, mixed integer linear programming, quadratic unconstraint binary optimisation, maximum satisfiability problem, federated machine learning, efficient pre-training of restricted boltzmann machines and deep neural networks, subset sum problems or integer factorisation.

Learn more about Dynex

Get DNX

Technological Deep-Dive

Join the Dynex Community

About

The Neuromorphic Computing Platform based on a new flexible blockchain protocol

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published