Overview
Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Explore a conference presentation introducing ClubHeap, a novel hardware-friendly heap data structure designed to overcome performance and scalability challenges in programmable packet scheduling for network devices. Learn how this innovative solution addresses limitations of existing PIFO (Push-In-First-Out) priority queue implementations by achieving theoretical lower bound performance with the ability to schedule a flow in every clock cycle. Discover the optimized hardware architecture that reduces circuit complexity while enabling higher clock frequencies, resulting in superior scheduling performance compared to state-of-the-art solutions. Examine the inherent scalability and flexibility advantages that make ClubHeap ideal for implementation in programmable switches and SmartNICs to support various scheduling algorithms. Review the FPGA-based hardware prototype development and comprehensive evaluation comparing ClubHeap against existing solutions, including its ability to provide graceful trade-offs between throughput and resource consumption through parameter adjustments for adaptability across different target devices.
Syllabus
NSDI '25 - ClubHeap: A High-Speed and Scalable Priority Queue for Programmable Packet Scheduling
Taught by
USENIX