Transforming Genomics with 5 PB Big Omics Data Engine Cray CS300-AC Supercomputer

描述：我们的基因组科学家和他们的外部合作者正处于他们研究的关键时刻：现在，而且在未来几年里，他们的发现将受到他们可以访问的计算和存储的数量和质量的限制。尽管西奈山为一个出色的超级计算和存储基础设施提供了资金，并有专业人员对其进行操作和维护，但它取得了巨大的成功，现在供应不足，对其最大的用户群体--基于基因组学的研究人员--产生了负面影响。这些研究人员的科学生产力在很大程度上受到了现有的限制 基础设施由于长时间排队等待，特别是在关键时期，再加上由于供应严重不足而花费在管理数据存储上的超长时间。随着20个私人投资机构、他们的24个项目及其55个外部合作机构计划的日益复杂的科学和数据查询的爆炸性增长，包括超过4300万美元的NIH资金，这些有限的资源严重限制了研究人员与更广泛的科学界合作和分享他们的数据的能力。为了使这些研究人员能够蓬勃发展，我们提出了一个“大数据引擎”(BODE)工具：一种专门的、专门的数据分析超级计算机，专为西奈山的特定科学需求而定制。这样的仪器将通过使用2484个英特尔哈斯韦尔企业内核将其吞吐量提高至少3倍到10倍。可用的存储空间将扩大3倍以上，达到5 PB，从而大大加强基于基因组学的一系列疾病类别的研究，包括自闭症、糖尿病患者的胰岛素抵抗、精神分裂症和相关的行为障碍、心脏护理、药物成瘾和抑郁的起源以及癌症进展。将基因组研究人员转移到一台新机器上还会产生第二个影响，即为我们的下一个需求最高的用户群体--结构和化学生物学研究人员--释放现有超级计算机上的计算周期和存储空间。为了认识到对这种仪器的强烈科学需求，西奈山已经同意资助高达400万美元 在运营成本方面。

DESCRIPTION: Our genomics-based scientists and their external collaborators are at a critical point in their research: now, and increasingly over the next years, their discoveries will be limited by the quantity and quality of computing and storage that they can access. While Mount Sinai has funded an excellent supercomputing and storage infrastructure along with professional staff to operate and maintain it, it has been a runaway success and it is now under-provisioned, negatively impacting its largest group of users the greatest: the genomics-based researchers. The scientific productivity of these researchers is greatly inhibited on this existing infrastructure due to long queue wait times especially during critical periods compounded by the extraordinary amount of time spent managing their data storage due to a substantial under-supply. With the explosion of progressively complex scientific and data queries planned by the 20 PIs, their 24 projects and their 55 external collaborating institutions comprising over $43 million in NIH funding, these limited resources are severely restricting the ability of the researchers to collaborate and to share their data with the broader scientific community. To enable these researchers instead to flourish, we propose a "Big Omics Data Engine" (BODE) instrument: a dedicated, specialized data analytic supercomputer customized for Mount Sinai's specific scientific needs. Such an instrument will increase their throughput by at least 3X and up to 10X by using 2,484 Intel Haswell enterprise cores. The available storage space will expand by over 3X to 5 petabytes, thus greatly enhancing the genomics-based research in a spectrum of disease categories including autism, insulin resistance in diabetics, schizophrenia and related behavioral disorders, cardiac care, the origins of drug addiction and depression, and cancer progression. Moving the genomics researchers to a new machine will also have the secondary effect of freeing up compute cycles and storage on our existing supercomputer for our next highest demand group of users: the structural and chemical biology researchers. In recognition of the intense scientific need for this instrument, Mount Sinai has agreed to fund up to $4 million in operating costs.