Discover more at https://dev.to/rawveg/when-20-watts-beats-20-megawatts-1np0
#HumanInTheLoop #NeuromorphicComputing #AIEnergyEfficiency #ComputationalArchitecture
Von-Neumann-Architektur ist nicht mehr alternativlos.
Forscher der Sandia Labs zeigen, dass neuromorphe Hardware komplexe PDEs effizient berechnet. Bisher galt das als Domäne klassischer HPC-Cluster. Durch ereignisbasierte Verarbeitung sinkt der Energiebedarf bei Diffusionsmodellen drastisch.
Ist Event-based Computing die einzige Lösung für das Energieproblem? #SandiaLabs #NeuromorphicComputing #HPC
https://www.all-ai.de/news/news26/intel-nvidia-neumann
There's a whole field trying to recreate Commander Data's brain from #StarTrek in real life. In #Neuromorphic computing they're trying to optimize computers that work in ways more like biological networks. 🧠+🖥=🤖
#TechnologyNews #NeuromorphicComputing #AI #Technology
https://www.theregister.com/2026/01/09/artificial_brains_supercomputer
Neuromorphs at work:
"Exploiting neuro-inspired dynamic sparsity for energy-efficient intelligent perception", Zhou et al. 2025 (Shih-Chii Liu and Tobi Delbruck's labs)
https://www.nature.com/articles/s41467-025-65387-7
A perspective paper on neuro-inspired vision on how to improve the energy efficiency of machine learning for perception, based on the observed sparse firing of biological neural circuits.
Edge AI enables intelligent perception in sensory devices, yet at excessive energy costs. This Perspective outlines a neuro-inspired vision for efficient edge perception, sketching the design space of data-driven and stateful dynamic sparsity to selectively activate sensors, memory, and compute.
First neuromorphic computer with 2 billion neurons
The system consists of 15 blade-type neuromorphic servers with 960 Darwin 3 chips. Each chip supports more than 2.35 million spiking neurons and hundreds of millions of synapses.
En Chine, une équipe de chercheurs a mis au point un ordinateur neuromorphique d'une puissance jamais atteinte. Cette innovation va permettre de concevoir des systèmes informatiques plus performants et plus autonomes.