Last week, we had the pleasure of featuring Dr. Anuj Pathania (Assistant Professor, University of Amsterdam) who delivered an engaging talk titled: “Edge Inference on HMPSOCs”
at Karlsruhe Institute of Technology (KIT) and Technische Universität Darmstadt , while promoting the COIN-3D project.

Dr. Anuj Pathania discussed how heterogeneous multi-processor systems-on-chip (CPU, GPU, NPU combinations on a single die) enable cloud-free, low-latency, energy-efficient AI inference at the edge. He highlighted both the opportunities and challenges of running AI workloads across multiple processors, especially when balancing performance, accuracy, power, and reliability.

🔍 His research connects strongly to COIN-3D’s mission of advancing sustainable, reliable, and low-power computing for emerging 3D architectures.

A big thank-you to Anuj Pathania for sharing his work with our community!