smartnic
Cora: Accelerating Stateful Network Applications with SmartNICs
Xi, Shaoke, Gao, Jiaqi, Liu, Mengqi, Cao, Jiamin, Li, Fuliang, Bu, Kai, Ren, Kui, Yu, Minlan, Cai, Dennis, Zhai, Ennan
With the growing performance requirements on networked applications, there is a new trend of offloading stateful network applications to SmartNICs to improve performance and reduce the total cost of ownership. However, offloading stateful network applications is non-trivial due to state operation complexity, state resource consumption, and the complicated relationship between traffic and state. Naively partitioning the program by state or traffic can result in a suboptimal partition plan with higher CPU usage or even packet drops. In this paper, we propose Cora, a compiler and runtime that offloads stateful network applications to SmartNIC-accelerated hosts. Cora compiler introduces an accurate performance model for each SmartNIC and employs an efficient compiling algorithm to search the offloading plan. Cora runtime can monitor traffic dynamics and adapt to minimize CPU usage. Cora is built atop Netronome Agilio and BlueField 2 SmartNICs. Our evaluation shows that for the same throughput target, Cora can propose partition plans saving up to 94.0% CPU cores, 1.9 times more than baseline solutions. Under the same resource constraint, Cora can accelerate network functions by 44.9%-82.3%. Cora runtime can adapt to traffic changes and keep CPU usage low.
- North America > United States > New York > New York County > New York City (0.04)
- North America > United States > Massachusetts > Suffolk County > Boston (0.04)
- North America > United States > Washington > King County > Renton (0.04)
- (7 more...)
- Telecommunications > Networks (0.46)
- Information Technology > Services (0.46)
- Information Technology > Artificial Intelligence > Machine Learning (1.00)
- Information Technology > Communications > Networks (0.93)
- Information Technology > Software > Programming Languages (0.93)
- Information Technology > Artificial Intelligence > Representation & Reasoning > Constraint-Based Reasoning (0.34)
Trends Continue to Push Processing to the Edge for AI
As connected devices proliferate, new ways of processing have come to the fore to accommodate device and data explosion. For years, organizations have moved toward centralized, off-site processing architecture in the cloud and away from on-premises data centers. Cloud computing enabled startups to innovate and expand their businesses without requiring huge capital outlays on data center infrastructure or ongoing costs for IT management. It enabled large organizations to scale quickly and stay agile by using on-demand resources. But as enterprises move toward more remote models, video-intensive communications and other processes, they need an edge computing architecture to accommodate data-hogging tasks.