How to Build Ultra-Efficient AI Inference for Edge Devices | Sam Fok (femtoAI)

At the Bridging Silicon Valley and Taiwan: Semiconductor & AI Synergies event, Sam Fok, Co-Founder & CEO of femtoAI, shares real-world insights on building ultra-efficient AI inference for embedded and edge devices. Drawing from hands-on experience at the intersection of silicon and software, Sam explains how breakthroughs in co-design are enabling powerful AI to run on constrained hardware—dramatically reducing energy consumption, latency, and cost.

In this talk, Sam breaks down why efficiency is becoming the defining constraint of the next AI wave, how startups should think about deploying intelligence into everyday physical products, and why Taiwan’s advanced semiconductor manufacturing ecosystem is critical to turning cutting-edge research into scalable, real-world systems. His perspective is especially relevant for founders working on AI hardware, edge computing, robotics, and next-generation embedded systems.

🚨 Reminder: Applications for IC Taiwan Grand Challenge (ICTGC) Batch 4 are now open, with the deadline coming up soon. Selected teams gain access to mentorship, corporate partners, manufacturing resources, and opportunities to scale with Taiwan’s world-class semiconductor ecosystem.

👉 Apply here: https://ictaiwanchallenge.org

For full event context and more talks from Bridging Silicon Valley and Taiwan: Semiconductor & AI Synergies, visit:

👉 https://www.sparknify.com/bridging-svt

Organized by Taiwan National Science & Technology Council

Implemented by Taipei Computer Association

Produced by Sparknify

#SamFok #femtoAI #EdgeAI #AIHardware #Semiconductors #DeepTech #StartupScaling #ICTGC #TaiwanTech #SiliconValley #EmbeddedAI #Sparknify