The Wisdom of Athena: Extending the scientific impact of the NSF's most cost-effective supercomputer

1041710科瓦奇这个奖项将资助雅典娜，166-TF克雷XT 4在田纳西大学？美国国家计算科学研究所（NICS），NSF的第三大和最具成本效益的超级计算机。每核心小时的运行成本仅为1美分多一点，这对研究人员来说是一个前所未有的机会。由于Athena的上级互连，在其他地方难以扩展的应用程序在Athena上非常成功。此外，将Athena与Kraken共同分配的能力将使两台机器的用户受益。Athena上可以运行长时间运行的、相对较小的核心数作业（60小时作业，最多18，048个核心），而较大的Kraken上则可以运行更短、更大的核心数作业（24小时作业，最多99，072个核心），从而使用户能够从每个平台的优势中受益。在过去的两年里，Athena一直在成功可靠地推动科学，无论是在TeraGrid的一般生产模式下，还是在气候和高能物理组的专用模式下。NICS建议继续运营Athena 12个月。

1041710KovatchThis award will fund Athena, the 166-TF Cray XT4 at University of Tennessee?s National Institute for Computational Sciences (NICS), NSF's third largest and most cost effective supercomputer. At just over one cent per core-hour to operate, it represents an unprecedented opportunity for researchers to perform science. Applications that have difficulty scaling elsewhere have been very successful on Athena due to its superior interconnect. In addition, the ability to co-allocate Athena with Kraken will benefit users of both machines. Long running, relatively small core count jobs (60 hour jobs up to 18,048 cores) can be run on Athena, leaving shorter, bigger core count jobs on the larger Kraken (24 hour jobs up to 99,072 cores), allowing users to benefit from the strengths of each platform. Athena has been enabling science successfully and reliably for the last two years, both in general production mode for the TeraGrid, and also in dedicated mode for climate and high energy physics groups. NICS proposes to continue operating Athena for 12 additional months.