Squash - 딥러닝 서비스 추론 속도 개선기
Explore a technical conference presentation from SK TECH SUMMIT 2023 focused on improving deep learning service inference speed through Squash, SK hynix's deep learning optimization technology brand. Learn how AI practitioners can effectively utilize expensive GPU resources through deep learning optimization techniques, particularly relevant in the era of generative AI and large language models. Discover how the AI Bartender team at SK hynix enhances service quality while reducing hardware costs through various optimization methods. The speaker, Hyunwoo Je from SK hynix, shares his journey from developing AI platforms and speakers at a telecommunications company to working on Digital Transformation initiatives at SK hynix, where he has developed conversational chatbots and internal education content recommendation systems, and now focuses on On Sensor AI and Efficient AI technologies.