Meet us at IEDM DEC 2022, San Francisco CA
Our work is rooted in more than a decade of research at Purdue University. It has expanded to UC Berkeley, UC Santa Barbara, and Tohoku University, There have been dozens of journal publications spanning physics and engineering, such as Nature Nanotechnology, Nature, Nature Electronics, Physical Review X, Applied Physics Review, Applied Physics Letters, IEEE Access, IEEE Spectrum, and Nano letters, to name a few.
Ludwig Computing is commercializing this technology.
Energy-Based Models in Hardware
Design of electronic and Spintronic devices whose intrinsic physics represents Bayesian Belief Networks, Ising Solvers, and Boltzmann machines.
A specially designed network of mixed-Signal intrinsically stochastic devices optimizes an objective function to factorize integers via a quantum-inspired algorithm
Massively parallel p-bits compute autonomously
The sequencer-less design allows autonomous computation resulting in massive parallelization and local computing alleviating the Von Neumann bottleneck.
Benchmarking a probabilistic co-processor
P-Computer architecture outperforms CPUs and GPUs for optimization, Bayesian inference, and Monte Carlo stochastic sampling