A Self-Designed Heterogeneous Accelerator for Exascale High Performance Computing
-
摘要: 高性能计算(high performance computing, HPC)是推动科学技术发展的基础性领域之一,当前,作为超级计算机系统“下一个明珠”的E级高性能计算时代已经来临.面向E级高性能计算的加速器领域成为了全球高端芯片的竞技场.国际上,AMD、英伟达和英特尔公司已经占据这一领域多年.作为国内最早开始自主处理器设计的优势单位之一,国防科技大学一直以来都是高性能加速器领域强有力的竞争者.主要对国防科技大学自主设计的面向E级高性能计算的加速器芯片进行介绍,该芯片采用了CPU+GPDSP的异构融合架构,具备高性能、高效能和高可编程性的特点,有望成为新一代E级超算系统的核心计算芯片.Abstract: High performance computing (HPC) is one of the basic fields to promote the development of science and technology. Exascale HPC era, recognized as “the next crown of supercomputer”, is coming. The accelerator field for exascale HPC has gradually developed into the arena of the most high-end chips in the world. The international famous companies, such as AMD,NVIDIA and Intel, have occupied this field for several years. As one of the organizations which independently designed processors in China, National University of Defense Technology (NUDT) has always been a strong competitor in HPC accelerator field. This paper introduces an accelerator for exascale HPC which is self-designed by NUDT. It adopts a heterogeneous architecture with CPU and general purpose digital signal processor (GPDSP). It has the characteristics of high performance, high efficiency and high programmability, and is expected to be the key computing chip of our new exascale supercomputer system.
-
-
期刊类型引用(3)
1. 陈晓江,杨晓奇,陈广豪,刘伍颖. 混合BERT和宽度学习的低时间复杂度短文本分类. 山东大学学报(工学版). 2024(04): 51-58+66 . 百度学术
2. 金泽熙,李磊,刘继. 基于改进领域分离网络的迁移学习模型. 计算机应用. 2023(08): 2382-2389 . 百度学术
3. 张娟飞. 基于VMD的数控机床旋转机械运行路径偏离故障检测方法. 机械与电子. 2022(11): 30-34 . 百度学术
其他类型引用(2)
计量
- 文章访问数: 939
- HTML全文浏览量: 6
- PDF下载量: 525
- 被引次数: 5