Peter Torelli - President, EEMBC
The sudden popularity of ADAS has driven a surge of demand for high-performance CPU and GPU hardware in automobiles, creating problems for Tier 1 manufacturers and OEMs.
- What is the best way to balance cost and design complexity against this enormous need for compute power never before seen in conventional cars?
- Discuss how to provide benchmarks and analysis tools for systems like these, and analyze the ubiquitous ADAS task: real-time image classification.
- Get to grips with image processing and CNN classification elements which can be arranged in a directed acyclic graph (DAG) to characterize the performance of heterogeneous architectures seen in many ADAS platforms