Daniel Klünder
Prof. Dr.-Ing. Daniel Klünder is a Professor of Electrical Engineering at the Baden-Wuerttemberg Cooperative State University (DHBW) Stuttgart. Before joining academia, he gathered extensive industry and leadership experience in the automotive and software sectors, focusing on embedded software and vehicle electronics in senior roles at ETAS, FEV, and Daimler. He holds a PhD in embedded software design from RWTH Aachen University.
His current work operates at the intersection of hardware architecture and strategic system design, with a strong focus on Next-Gen Edge-AI for autonomous systems in defense and critical infrastructure. He specializes in overcoming strict SWaP-C (Size, Weight, Power, and Cost) constraints through energy-efficient "Zero-Overhead" architectures.
Session
Deploying high-performance AI inference on autonomous drones requires a precise balance between computational throughput and a strict 1W power envelope. This paper presents a vertical design space exploration (DSE) of the RISC-V Gemmini accelerator, scaling from 8x8 to 32x32 mesh configurations in the SkyWater 130nm process. Through an end-to-end evaluation using a YOLOv4-tiny model on the VisDrone dataset, we demonstrate a 74.75% reduction in model memory footprint via INT8 quantization and a speedup of up to 2352x compared to a RISC-V CPU baseline. Our results indicate that while the 32x32 mesh excels in peak throughput, the 16x16 mesh represents the optimal “sweet spot” for 1W-limited drone chiplets, combining high performance with manageable leakage and area.