학사
정기세미나
View
SAIT의 수퍼컴을 활용한 Hyperscale AI와 Machine Learning Force Fields 연구 현황 | ||
---|---|---|
담당자 김용덕 박사(삼성전자) | 세미나 일자 2023.12.13 Wed | 조회수 470 |
1. Education
l B.S in Computer Science and Engineering, POSTECH, 2005
l Ph.D in Computer Science and Engineering, POSTECH, 2015
2. Working Experience in Samsung
Ongoing Projects:
- Researching on machine learning inter-atomic potential for accelerating quantum-mechanical atomistic simulations. (2021~present)
Visiting Scholar @ CMU (2019~2020)
Completed Projects:
- Building Hyper-scale AI R&D infra in Samsung Super Computer. Reported outstanding BERT training performance using 1024 A100 GPUs to MLPerf Training benchmark v1.1, v2.0 and v2.1. (2020~2022)
- (2015~2018) Mobile GPU-based deep learning inference framework development and commercialization
- (2017~2018) Deep learning based automatic SEM images classification methodology development and application to Through-Silicon Via (TSV) process
3. Publication:
https://scholar.google.com/citations?user=GmXJVx4AAAAJ&hl=ko