OptimusPrime: Unleash Dataplane Programmability through a Transformable Architecture
Network dataplane calls for better programmability. Current programmable network
processing chips are based on either pipeline or multi-core Run-To-Completion (RTC) …
processing chips are based on either pipeline or multi-core Run-To-Completion (RTC) …
hyDNS: Acceleration of DNS Through Kernel Space Resolution
J Bardinelli, Y Zhang, J Su, L Huang, A Parilla… - Proceedings of the …, 2024 - dl.acm.org
The Domain Name System (DNS) is a core component of Internet infrastructure, mapping
domain names to IP addresses. The recursive resolver plays a critical role in this process …
domain names to IP addresses. The recursive resolver plays a critical role in this process …
BOAD: Optimizing Distributed Communication with In-Kernel Broadcast and Aggregation
J Su, Y Zhang, L Huang, W Zhang - … 2024 Workshop on eBPF and Kernel …, 2024 - dl.acm.org
Efficient communication is crucial for the performance of big data and distributed computing
systems. Two key communication patterns in these systems are broadcasting, which …
systems. Two key communication patterns in these systems are broadcasting, which …
PINCH: Accelerating Distributed GNN Training through In-Kernel Operation Using eBPF
J Su, Y Zhang, W Zhang - Machine Learning for Computer Architecture and … - openreview.net
Graph Neural Networks (GNNs) have achieved state-of-the-art performance on various
graph-related tasks, but training GNNs on large-scale graphs remains challenging due to …
graph-related tasks, but training GNNs on large-scale graphs remains challenging due to …