AI automated 2.4 GHz CMOS low-noise amplifier optimization cuts power 62%, linearity improves signal, increases valid simulation rates

https://www.eurekalert.org/news-releases/1139943

"AI search random candidate circuits frequently violate strict foundry constraints... overcome using validity gate screening simulated candidates during Bayesian optimization, raising buildable circuit simulation success from 47.5 to 95.7%, pre-teaches algorithm resonance/ impedance-matching creating valid warm-starts... power reduced 62% from 14.1 mW to 5.4 mW, 10.7 dB enhanced signal linearity... improved secondary amplifier... analog RF chip AI assistants in: smartphones, Wi-Fi routers, wearables"