[CLSTR]
< Back to situations

Monitor this situation.

[SITUATION] · [QUIET] · [TECHNOLOGY]

2 clusters · 2 sources · 8 days · First seen · Last updated

South Korean neuromorphic semiconductor research

Overview

South Korean researchers are advancing neuromorphic semiconductor technology designed to mimic biological neurons for low-power AI applications.

Initial developments at KAIST focused on a ‘Programmable Probabilistic Neuron’ (PPN) that utilizes natural semiconductor noise to process information. By using memristor devices to convert electrical noise into probabilistic spikes, the technology achieved high accuracy in both motion recognition and speech recognition tasks.

Subsequent research by GIST and KENTECH addressed stability issues in oxide-based memristors. By employing the Marangoni effect to self-assemble cobalt oxide nanoparticles into a precise checkerboard pattern with 3nm nanogaps, researchers created a volatile memristor that ensures consistent electrical current paths and high stability at low voltages.

Entities

KAIST · KENTECH · GIST · Kyung-min Kim

Timeline

  1. 20 days ago

    [TECHNOLOGY] 2 sources
    GIST and KENTECH develop brain-inspired AI memristor

    GIST and KENTECH researchers developed a stable, low-power volatile memristor using self-assembled cobalt oxide nanoparticles to mimic biological neurons for neuromorphic computing.

  2. 27 days ago

    [TECHNOLOGY] 2 sources
    KAIST researchers develop brain-inspired neuromorphic semiconductor

    KAIST researchers developed a neuromorphic semiconductor that uses electrical noise to mimic brain neurons, achieving high accuracy in both motion and speech recognition for low-power AI.

Sources

biz.heraldcorp.com · jnilbo.com