RDMA Courses and Certifications

USENIX

ShiftLock: Mitigate One-sided RDMA Lock Contention via Handover

Explore ShiftLock, an innovative RDMA lock mechanism that reduces contention through direct client-to-client lock handover, improving performance and efficiency in distributed storage systems.

• YouTube
• 15 minutes
• On-Demand
• Free Video

SNIAVideo

Everything You Wanted to Know About RDMA But Were Too Proud to Ask

Explore RDMA technology's fundamentals, benefits, and implementations including RoCE and InfiniBand, with insights into its historical development, components, and crucial role in modern high-performance computing environments.

• YouTube
• 1 hour
• On-Demand
• Free Video

Anyscale

RDMA P2P Deep Dive - KvCache Transfer, Weight Updates and MoE Routing at Perplexity

Explore RDMA peer-to-peer communication for LLMs, covering KvCache transfer, weight updates, and MoE routing to overcome collective communication bottlenecks in scalable AI systems.

• YouTube
• 31 minutes
• On-Demand
• Free Video

Anyscale

Ray Direct Transport - RDMA Support in Ray Core

Discover how Ray Direct Transport eliminates GPU bottlenecks by enabling zero-copy, direct device-to-device data transfer, dramatically accelerating distributed AI workloads.

• YouTube
• 32 minutes
• On-Demand
• Free Video

SNIAVideo

RDMA Tech in a Nutshell

Discover RDMA technology fundamentals and operational mechanics through expert insights from Marvell's networking specialist in this concise technical overview.

• YouTube
• 11 minutes
• On-Demand
• Free Video

USENIX

Achieving Wire-Latency Storage Systems by Exploiting Hardware ACKs

Discover Juneberry, a low-latency communication framework that achieves wire-speed storage by exploiting hardware ACKs for single round-trip commits with CPU bypass.

• YouTube
• 16 minutes
• On-Demand
• Free Video

USENIX

Cuckoo for Clients - Disaggregated Cuckoo Hashing

Discover RCuckoo, a disaggregated key/value store using cuckoo hashing and one-sided RDMA operations for high-performance data access with enhanced locality and fault tolerance.

• YouTube
• 18 minutes
• On-Demand
• Free Video

DevConf

SMC-R Protocol in Storage Software

Explore SMC-R and SMC-D protocols for RDMA communications, their implementation in software-defined storage, performance comparisons, and practical application integration.

• YouTube
• 15 minutes
• On-Demand
• Free Video

Open Compute Project

Powering Next Generation Switch Architecture for AI Hyperscale Infrastructure through Open Standards

Discover purpose-built network switch hardware designed for AI hyperscale data centers, featuring ultra-high bandwidth, RDMA support, and optimized fabric scheduling for enhanced performance.

• YouTube
• 24 minutes
• On-Demand
• Free Video

Open Compute Project

Enable IBGDA Support in a GPU Agnostic Manner for Open AI Systems

Discover how to extend IBGDA technology beyond Nvidia GPUs to work with any GPU and Ethernet networks, optimizing RDMA performance for open AI systems using RoCE.

• YouTube
• 13 minutes
• On-Demand
• Free Video

Tech Field Day

Reliable Transport of Lossy Fabrics with Falcon - A Low Latency Hardware Solution

Discover how Falcon enables reliable RDMA messaging and NVMe commands across cloud networks, optimizing performance for AI/ML, HPC, and storage workloads with low-latency transport.

• YouTube
• 37 minutes
• On-Demand
• Free Video

Oracle

Building a High Performance Network in the Public Cloud Using RDMA - First Principles

Explore Oracle Cloud Infrastructure's high-performance networking using RDMA, covering challenges, optimizations, and unique features for low-latency workloads in public cloud environments.

• YouTube
• 40 minutes
• On-Demand
• Free Video

Donato Capitella

Low-Latency Strix Halo Cluster with RDMA - RoCE/Intel E810 and vLLM, Framework Desktop Boards

Explore building a 2-node Strix Halo cluster with RDMA networking and vLLM tensor parallelism, including hardware setup, custom cooling, and ROCm patches for high-performance inference.

• YouTube
• 16 minutes
• On-Demand
• Free Video

XDP/BPF in RDMA - RXE Implementation

1 review

Explore XDP/BPF integration in RDMA(RXE) for enhanced network performance and programmability. Learn about cutting-edge techniques in Linux networking and data processing.

• YouTube
• 29 minutes
• On-Demand
• Free Video

USENIX

Flor - An Open High Performance RDMA Framework Over Heterogeneous RNICs

Explore Flor, an innovative framework enabling high-performance RDMA across diverse network interface cards, addressing bandwidth imbalances and enhancing datacenter application efficiency.

• YouTube
• 15 minutes
• On-Demand
• Conference Talk