NuAxi
NuAxiSemiconductor trade intelligence

Edge Revolution: Quadric’s GPNPU Makes On‑Device AI Practical

What happened

EEJournal's podcast (episode 708, air date November 13, 2026) interviews Quadric CEO Veerbhan Kheterpal about the company's fully programmable Chimera GPNPU (general-purpose neural processing unit) architecture for AI inference. Topics include combining neural network acceleration with conventional computing, growing demand for on-device AI token generation, challenges of running complex AI models on power-constrained devices, applications such as automatic braking in cars, and a trend of enterprises weighing fixed hardware investment against continuous token OpEx for local AI deployment. This is background technical interview content with no component price movement, capacity, inventory or lead-time data, and the source does not state the cause or any specific supply impact.

Impact and signal

Low impactImpact score: 15/100

Affected components and suppliers

Chimera GPNPU
Quadric
edge AIGPNPUAI inferenceon-device AIpodcast interviewprocessor architecture

Source and time

Source: EE Journal · 11/20/2026 · View original source

Open the NuAxi live feed →

Real-time pricing, supply signals and AI analysis for semiconductor traders