CNS Core: Medium: SmartNIC-Accelerated Database Systems

数据库为世界上许多最大和最常用的Internet服务和应用程序提供动力。今天，这些系统以分布式方式跨多台机器甚至跨数据中心运行，这是由不断增长的数据量和不断增长的Internet服务规模所驱动的。网络通信已成为分布式数据库的一个严重瓶颈。近年来，人们发明了一种新的技术——智能网卡（SmartNIC），它将智能添加到网卡（NIC）中，以减少主机处理器的计算开销。这种新的硬件设备在加速分布式数据库操作方面具有很大的潜力。这个项目的目标是探索这个设计领域的机会。拟议的工作将为在线分析处理（OLAP）和在线事务处理（OLTP）工作负载开发smartnic加速的高性能分布式数据库。这项工作可以改善内部部署和云分布式数据库，并为整个社会增加重大价值。提案工作可以潜在地集成到商业分布式数据库中，从而为互联网服务和应用程序带来高性能和低延迟，从而为个人客户带来好处。通过该项目开发的软件将被整合到现有的开源大数据栈和数据库系统中。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Databases power many of the largest and most commonly used Internet services and applications in the world. Today, these systems are run in a distributed fashion across multiple machines or even across data centers, driven by the growing volumes of data and the growing scale of Internet services. The network communication is becoming a severe bottleneck in distributed databases. Recently, a new technology, SmartNIC, has been invented to add intelligence into the network interface card (NIC) to reduce the computation overhead of host processor. This new hardware device has great potential in accelerating distributed database operations. The goal of this project is to explore opportunities in this design space. The proposed work will develop SmartNIC-accelerated high performance distributed databases for both online-analytical processing (OLAP) and online-transactional processing (OLTP) workloads. The work can improve both on-premise and cloud distributed databases and add significant value to society at large. The proposal work can potentially be integrated into commercial distributed databases, leading to high performance and low latency for internet services and applications which can bring benefits to individual customers. The software developed through this project will be incorporated into existing open source big data stacks and database systems.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.