Open Compute Project Courses

Beyond 400Gbps Lane Rate Transmission

Discover the feasibility of optical and electrical links at 400Gbps per lane net bit rate, a technology set to power future 3.2Terabit networking using conventional modulation techniques.

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

Fitting Big Quantum Computers Through Small Doors

Discover the challenges and solutions for installing large quantum computing systems in existing facilities with space constraints.

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

Enhancing RAS in AI Hardware and High-Performance Computing with Real-Time Health Monitoring

Discover how Real-Time Health Monitoring enhances reliability in AI hardware and data centers by predicting failures, preventing downtime, and mitigating silent data corruption through embedded deep data monitoring.

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

System Integration of a Computing Cluster - Heat Recovery and Adsorption Chiller for Cold Production

Discover how Qarnot Computing reuses server heat for cold production through an innovative system coupling Computing Boilers with silica gel adsorption chillers, achieving impressive efficiency metrics and environmental benefits.

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

Unified Platform Configuration: OPF Initiative to Simplify Platform Configuration

Explore the Unified Platform Configuration initiative by OPF aimed at simplifying configuration processes for heterogeneous compute environments and modular hardware systems.

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

Software Infrastructure for Quantum Computer Integration with HPC

Explore the integration of quantum computing with high-performance computing infrastructure in this talk by Venkatesh Kannan from the Irish Centre of High-End Computing.

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

Nano-dielectric Oils for Immersion Cooled Data Centers

Explore how nano-dielectric oils enhance immersion cooling for data centers, improving thermal conductivity and enabling better heat removal for AI and HPC workloads.

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

State of the European HPC Market: Trends, Challenges and Opportunities

Explore the European HPC market landscape with insights on infrastructure developments, EuroHPC funding, AI Factories, and open hardware opportunities and challenges.

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

Enabling Scalable AI with 1.6T Pluggable Optics Based on 224G/lane

Explore how 1.6T pluggable optics with 224G/lane technology enables scalable AI infrastructure through emerging standards, power-efficient DSPs, and linear optics like LPO and LRO.

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

Practical Quantum Acceleration for HPC and AI

Discover how photonic quantum computing can accelerate HPC and AI workloads through practical applications and advantages.

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

AI Infrastructure Rack Management: Exploring Scalable Solutions Through Open-source Collaboration

Explore how AI infrastructure transforms data center management, requiring new open-source solutions beyond current designs to address workload complexities and system interdependencies for more resilient and responsive infrastructure.

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

Photonic Integrated Switch Fabrics: From Homogeneous to Heterogeneous Integrations

Explore large port count optical switches using photonic integrated circuits, from foundry fabrication to innovative heterogeneous integration methods like micro-transfer printing and additive manufacturing.

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

High-Performance Two-Phase Cooling under Different Cold Plate Orientations

Explore two-phase cooling for data centers, examining cold plate performance in various orientations including vertical blades and upside-down servers, demonstrating consistent efficiency without design modifications.

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

Compression Enabled MRAM Memory Chiplet Subsystems for LLM Inference Accelerators

Explore how compression-enabled MRAM memory chiplet subsystems can optimize LLM inference accelerators, addressing memory-bound challenges in GPU architectures.

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

Improving Performance of AI Cluster by Tuning AI Ethernet Switch Features in SONiC

Discover how to optimize AI cluster performance by fine-tuning Ethernet switch features like RoCEv2, ECN, PFC, and DLB in SONiC for minimized latency and maximized throughput.

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