作者: Scott Klasky , Matthew Wolf , Mark Ainsworth , Chuck Atkins , Jong Choi
关键词: Data science 、 Development plan 、 Scalability 、 Software 、 SPARK (programming language) 、 Big data 、 Computer science 、 Workflow 、 Data visualization 、 Data modeling
摘要: One of the core issues across computer and computational science today is adapting to, managing, learning from influx "Big Data". In commercial space, this problem has led to a huge investment in new technologies capabilities that are well adapted dealing with sorts human-generated logs, videos, texts, other large-data artifacts processed resulted an explosion useful platforms languages (Hadoop, Spark, Pandas, etc.). However, translating work enterprise space HPC community proven somewhat difficult, part because some fundamental differences type scale data timescales surrounding its generation use. We describe forward-looking research development plan which centers around concept making Input/Output (I/O) intelligent for users scientific community, whether they accessing scalable storage or performing situ workflow tasks. Much our based on experience Adaptable I/O System (ADIOS 1.X), next version software ADIOS 2.X [1].