Journal provides a multidisciplinary forum for scientists, researchers and engineers involved in research and design of nonlinear processes and phenomena, including the nonlinear modelling of phenomen...

YouTube video by Knut Jägersberg

arXiv abstract link

Adel Ouannas/Amina Aicha Khennaoui/Omar Naifar

While there is an abundance of research on neural networks that are “inspired” by the brain, few mimic the critical temporal compute features that allow the ...

Efficient processing of high-dimensional datasets demands minimal information loss and reduced computational complexity. Ding et al. report a memristor-based system that boosts data simplification eff...

Efficient processing of high-dimensional datasets demands minimal information loss and reduced computational complexity. Ding et al. report a memristor-based system that boosts data simplification eff...

Efficient processing of high-dimensional datasets demands minimal information loss and reduced computational complexity. Ding et al. report a memristor-based system that boosts data simplification eff...

SVD decomposes the data matrix \({{{\boldsymbol{X}}}}\) into three matrices: \({{{\boldsymbol{U}}}}\), formed by left singular vectors (LSVs); \({{{\boldsymbol{S}}}}\), holding the singular values; and \({{{\boldsymbol{V}}}}\), containing the right singular vectors (RSVs). MSVD employs a power iteration-deflation approach to sequentially extract these components. As illustrated in Fig. 2a, the computation for each singular vector is executed through an iterative loop (marked in blue). Starting with a randomly initialized input voltage vector (\(\widetilde{{{{\boldsymbol{v}}}}}\)), the process performs dual matrix multiplications on two memristor subarrays: one storing the data matrix (\({{{\boldsymbol{X}}}}\)) and the other storing the previously solved RSVs (\({{{\boldsymbol{V}}}}\)). The multiplication on the data matrix executes one step of power iteration, while the multiplications on previously solved RSVs calculates the components to be deflated. The multiplication output from RSVs is then subtracted from the output of the data matrix to update \(\widetilde{{{{\boldsymbol{v}}}}}\), effectively completing the deflation to isolate the target singular vector. These operations form one iteration step of the loop. Once the loop converges or stops, the newly solved RSV (\({{{\boldsymbol{v}}}}\), obtained from converged \(\widetilde{{{{\boldsymbol{v}}}}}\)) is updated into \({{{\boldsymbol{V}}}}\) for subsequent power iteration and deflation steps (illustrated in red loop in Fig. 2a). The corresponding singular value (\(s\)) and LSV (\({{{\boldsymbol{u}}}}\)) are then derived...

Kaspar, C., Ravoo, B. J., van der Wiel, W. G., Wegner, S. V. & Pernice, W. H. P. The rise of intelligent matter. Nature 594, 345–355 (2021). Laydevant, J., Wright, L. G., Wang, T. & McMahon, P. L. The hardware is the software. Neuron 112, 180–183 (2024). Stieg, A. Z. et al. Emergent criticality in complex turing b-type atomic switch networks. Adv. Mater. 24, 286–293 (2011). Avizienis, A. V. et al. Morphological transitions from dendrites to nanowires in the electroless deposition of silver. Cryst. Growth Des. 13, 465–469 (2013). Demis, E. C. et al. Atomic switch networks — nanoarchitectonic design of a complex system for natural computing. Nanotechnology 26, 204003 (2015). Kim, T. H. et al. Nanoparticle assemblies as memristors. Nano Lett. 9, 2229–2233 (2009). Sandouk, E. J., Gimzewski, J. K. & Stieg, A. Z. Multistate resistive switching in silver nanoparticle films. Sci. Technol. Adv. Mater. 16, 045004 (2015). Mallinson, J. B. et al. Avalanches and criticality in self-organized nanoscale networks. Sci. Adv. https://doi.org/10.1126/sciadv.aaw8438 (2019). Stieg, A. Z. et al. in Memristor Networks (eds Adamatzky, A. & Chua, L.) https://doi.org/10.1007/978-3-319-02630-5_10 (Springer, 2014). Milano, G., Michieletti, F., Pilati, D., Ricciardi, C. & Miranda, E. Self-organizing neuromorphic nanowire networks as stochastic dynamical...

The energy and latency budgets of edge artificial intelligence (AI) are far tighter than in data-centre computing, yet the workloads are no less demanding. Conventional von Neumann hardware spends ...