Open Compute Project Courses

AI Factory Cyber Security and Data Protection Through Enhanced IT-OT Resilience Chain Integration

Explore enhanced IT-OT resilience in AI factories, focusing on cybersecurity challenges in cooling and power equipment integration, and addressing vulnerabilities in systems lacking secure industrial IoT connections.

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

Privacy for AI from NP-Hard Problems - Universal Compute on Encrypted Data

Explore cryptographic methods for private AI, including MultiParty Computation, Threshold Cryptography, and Fully Homomorphic Encryption, with focus on challenges in evaluating operations like softmax in the encrypted domain.

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

Challenges and Trends in Sparse Matrix Multiplication on HPC Workloads

Explore recent advancements in sparse matrix multiplication for HPC, examining hardware accelerators, software innovations, and optimization techniques to enhance efficiency in scientific computing and AI applications.

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

LLM Inference Performance Projection

Explore MESA, an AI inference projection tool for evaluating LLM performance across different hardware and models, with detailed operation breakdowns and context length analysis.

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

A Blueprint for Scalable AI Computational Infrastructure

Discover a blueprint for building scalable AI computational infrastructure, presented by OCP's VP of Market Intelligence and Innovation, James Kelly.

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

AI System Validation - Meta Perspective: The Challenges of Debugging at Scale

Explore the challenges of debugging at scale in Meta's complex AI systems and learn about their approach to monitoring, automated testing, and collaborative debugging for effective issue resolution.

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

Extreme Savings through Modern Storage Architecture

Discover how modern storage architecture can slash processing resources by 50% and storage costs by 75% while enhancing AI-driven intelligence through unified data management and advanced reduction techniques.

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

OOP/FP GPU Programming and Debugging Using Open-Source GpuScript

Explore GpuScript, an open-source GPU programming language offering OOP/FP features and debugging capabilities for complex GPU computation, achieving benchmarked speeds up to 23 PFLOPS on a single GPU.

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

Mitigating Silent Data Corruption: Industry-Academia Collaboration and Progress

Explore the collaborative efforts between major tech companies and universities to address silent data corruption through the Server Compute Resiliency Specification, with insights on progress and future GPU-focused developments.

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

Meta's ORv3 "HPR Next" Ecosystem Solution for High Power AI Architectures

Explore Meta's development of Open Rack v3 (ORv3) High Power Rack solutions for AI architectures, including HPR v2, v3, and v4 designs, power supplies, backup units, and power transmission methods.

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

When Single-Phase Immersion Cooling is the Right Choice for EMEA

Explore the key considerations for implementing single-phase immersion cooling in EMEA data centers, including industry concerns and OCP's solutions.

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

Powering Sustainable HPC: Scotland's Advantage in the Next Generation of Data Centers

Discover how Scotland is addressing sustainability challenges in HPC data centers, exploring customer requirements and innovative solutions for the next generation of computing infrastructure.

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

Sustainable AI-Ready Capacity - Myth or Reality?

Explore how colocation data centers can scale sustainably for AI workloads through Conapto's OCP Ready V2 certified facility, featuring energy-efficient design and heat reuse innovations.

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

Accelerating Deployment Efficiency: Collaborative Innovations for Speed and Quality in Next-Gen Data Centers

Discover how Cloudflare and MiTAC's partnership revolutionizes hyperscale deployment, achieving 40% faster timelines through OCP-inspired hardware integration and optimized operational workflows for next-generation data centers.

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

Streamlining OCP Deployments with Integrated Rack Solutions

Discover how integrated rack solutions simplify OCP deployments with pre-designed, pre-validated systems that address hosting, power, and cooling needs while reducing integration challenges and accelerating time-to-market.

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