Race to Exascale
Race to Exascale
复制标题
竞相达到百兆亿级
DOI：
--
复制
发表时间：
2019
期刊：
影响因子：
--
通讯作者：
W. Kramer
中科院分区：
文献类型：
--
作者：
J. Dongarra;S. Gottlieb;W. Kramer
无论是领导力计算，旗舰计算还是普通的Exascale，在接下来的几年中，全世界的政府都计划在少数新的计算机系统上花费超过100亿美元，这些系统将努力达到这些绩效系统和项目反映了宽度和扩展的认识，即几乎所有的科学和工程努力现在都对计算能力的本质负责，不仅用于建模和模拟，而且用于数据分析，数据和机器学习。
Whether called leadership computing, flagship computing, or just plain exascale, over the next few years, governments around the world are planning to spend over 10 billion dollars on a handful of new computer systems that will strive to reach an exascale level of performance. These systems and projects reflect the widespread and expanding recognition that almost all science and engineering endeavors now are intrinsically reliant on computing power not just for modeling and simulation but for data analysis, big data, and machine learning. Scientists and engineers consider computers as “universal instruments” of insight.