USENIX Courses

Multitenant In-Network Acceleration with SwitchVM

SwitchVM: A practical approach for multitenant in-network acceleration on programmable switches, enabling cloud users to deploy custom data-plane filters with isolation, state management, and dynamic chaining capabilities.

• YouTube
• 17 minutes
• Self-Paced
• Free Video

Accelerating Skewed Workloads with Performance Multipliers in TurboDB Distributed Database

Innovative hybrid database architecture combining distributed and single-machine systems to significantly boost performance for skewed workloads, featuring novel concurrency control and replication techniques.

• YouTube
• 16 minutes
• Self-Paced
• Free Video

Understanding Routable PCIe Performance for Composable Infrastructures

Exploring routable PCIe fabric performance in composable infrastructures using rPCIeBench. Analyzes communication characteristics, traffic orchestration, and develops insights for optimizing flow scheduling in shared fabrics.

• YouTube
• 17 minutes
• Self-Paced
• Free Video

Passengers' Safety Matters: Experiences of Deploying a Large-Scale Indoor Delivery Monitoring System

Innovative system for monitoring indoor deliveries, enhancing safety in subway stations. Uses sensors and data analysis to detect violations, significantly reducing incident rates over three years.

• YouTube
• 17 minutes
• Self-Paced
• Free Video

iStack: A General and Stateful Name-based Protocol Stack for Named Data Networking

Innovative protocol stack for Named Data Networking, integrating ICN into OS kernels with efficient name resolution, interface abstraction, and socket compatibility, enhancing performance and resource management.

• YouTube
• 18 minutes
• Self-Paced
• Free Video

Brain-on-Switch: Advanced Intelligent Network Data Plane via NN-Driven Traffic Analysis

Innovative approach to enable neural network-driven traffic analysis at line-speed in programmable networks, overcoming limitations of tree-based models and enhancing intelligent network data plane capabilities.

• YouTube
• 14 minutes
• Self-Paced
• Free Video

Precise Data Center Traffic Engineering with Constrained Hardware Resources

Optimizing data center traffic routing with limited hardware resources, adapting allocation based on volume and precision impact, resulting in improved load balancing and reduced congestion.

• YouTube
• 17 minutes
• Self-Paced
• Free Video

Alea-BFT: Practical Asynchronous Byzantine Fault Tolerance

Explore Alea-BFT, an efficient asynchronous Byzantine Fault Tolerance protocol gaining adoption in Ethereum. Learn its simple design, performance benefits, and practical applications in cryptocurrency ecosystems.

• YouTube
• 15 minutes
• Self-Paced
• Free Video

Harvesting Idle Memory for Application-managed Soft State with Midas

Explores Midas, a system for efficiently managing soft state in applications using idle memory, improving performance while ensuring safety and responsiveness to memory pressure.

• YouTube
• 18 minutes
• Self-Paced
• Free Video

Sirius: Composing Network Function Chains into P4-Capable Edge Gateways

Automated system for composing network function chains into P4 programs for edge gateways, optimizing traffic processing and reducing programmer workload in cloud environments.

• YouTube
• 16 minutes
• Self-Paced
• Free Video

Reverie: Low Pass Filter-Based Switch Buffer Sharing for Datacenters with RDMA and TCP Traffic

Innovative buffer-sharing scheme for datacenters, improving RDMA and TCP performance through unified admission control and low-pass filtering, enhancing burst absorption and flow completion times.

• YouTube
• 15 minutes
• Self-Paced
• Free Video

Load Balancing with PReQuaL: Minimizing Latency in Distributed Systems

Innovative load balancing technique PReQuaL minimizes latency in distributed systems by probing server load and estimating latency, improving performance for YouTube and other Google services.

• YouTube
• 19 minutes
• Self-Paced
• Free Video

Near-realtime Earth Observation Via Query Bifurcation in Serval

Innovative system enabling near-realtime Earth observation insights by optimizing satellite and ground station computation, reducing latency from days to minutes for critical applications like disaster monitoring.

• YouTube
• 17 minutes
• Self-Paced
• Free Video

UFO: The Ultimate QoS-Aware Core Management for Virtualized and Oversubscribed Public Clouds

Optimizing multi-tenancy, virtualization, and resource oversubscription in public clouds through QoS-aware core management, addressing challenges in latency-critical workloads and VM resource contention.

• YouTube
• 15 minutes
• Self-Paced
• Free Video

A High-Performance Design, Implementation, Deployment, and Evaluation of The Slim Fly Network

Explore the first real-world Slim Fly network installation, offering cost and power advantages over traditional topologies. Learn about deployment, management, and performance analysis of this innovative low-diameter network.

• YouTube
• 19 minutes
• Self-Paced
• Free Video