GPU平台加速深度学习和高性能计算

2017.12.12

投稿:刘 华部门:计算机工程与科学学院浏览次数:

活动信息

时间: 2017年12月14日 13:30

地点: 校本部东区计算机大楼402室

报 告 人:付庆平 高级系统架构师 (英伟达)

报告时间:1214日(周四)13:3016:30

报告地点:校本部东区计算机大楼402

邀 请 人:沈文枫 副研究员

报告摘要:

NVIDIA is a company built upon great minds and groundbreaking research. With more than 120 scientists around the globe, the areas of focus include AI, self-driving cars, high-performance computing, graphics, VR, and augmented reality. Our researchers operate as a think tank, accelerating advances in product technologies, including AI innovations like facial animations and light transport.

The report will focus on Nvidia’s GPU Platform latest technology and product in HPC and AI. Topics will cover volta100 GPU, Three Ways to Accelerate Applications, especially on OpenACC programming method for easy accelerate your own HPC applications.

报告人简介:

付庆平,Nvidia资深技术顾问。长期从事高性能计算集群技术相关工作,曾参与天河二号超级计算机调试工作、负责商汤科技深度学习计算集群设计与调优工作等。目前在Nvidia主要从事HPC及深度学习系统架构工作。