谷歌浏览器插件
订阅小程序
在清言上使用

Trillion Parameter AI Serving Infrastructure for Scientific Discovery: A Survey and Vision

PROCEEDINGS OF THE IEEE/ACM 10TH INTERNATIONAL CONFERENCE ON BIG DATA COMPUTING, APPLICATIONS AND TECHNOLOGIES, BDCAT 2023(2023)

引用 0|浏览39
暂无评分
摘要
Deep learning methods are transforming research, enabling new techniques, andultimately leading to new discoveries. As the demand for more capable AI modelscontinues to grow, we are now entering an era of Trillion Parameter Models(TPM), or models with more than a trillion parameters – such as Huawei'sPanGu-Σ. We describe a vision for the ecosystem of TPM users andproviders that caters to the specific needs of the scientific community. Wethen outline the significant technical challenges and open problems in systemdesign for serving TPMs to enable scientific research and discovery.Specifically, we describe the requirements of a comprehensive software stackand interfaces to support the diverse and flexible requirements of researchers.
更多
查看译文
关键词
Artificial Intelligence,Grid Computing,Deep Learning Applications,Systems Design,Survey
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要