Learn about optimizing synaptic weight distribution for Analog Computation-in-Memory (ACIM) devices in this technical presentation from SK TECH SUMMIT 2022. Explore the core technologies needed to implement energy-efficient 'brain-like' hardware for the Post von Neumann era. Discover how Memory-Centric NVM technology combined with ACIM serves as a crucial component for AI edge devices performing object recognition and pattern classification. Follow SK Hynix's Team Leader Park Sang-soo from the Future Memory Research Team as he presents technical approaches for optimizing synaptic weights and discusses this seeding technology's potential for creating new business models. Gain insights into the development of convergent semiconductor technology where memory semiconductors can independently perform both computation and processing, with a special focus on neuromorphic computing possibilities.
Overview
Syllabus
[SK TECH SUMMIT 2022] Analog Computation-in-Memory 향 시냅스 소자의 가중치 산포 최적화 연구
Taught by
SK AI SUMMIT 2024