Memory Courses

Open Compute Project

Co-Designing for Scale - CXL Based Memory Solution for Data Centric Workloads

Explore CXL-based memory solutions for AI workloads, featuring disaggregated architectures, composable systems, and open-source frameworks to optimize performance and reduce costs.

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

Open Compute Project

Memory for Every AI Need - Samsung's Next-Gen Memory Solutions to Lead the Future of AI

Discover Samsung's next-generation memory solutions designed for AI workloads, including HBM, MRDIMM, CAMM, and CXL technologies to enhance bandwidth, capacity, and power efficiency.

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

Open Compute Project

Memory Technology Optimized for At-Scale AI Systems - Bandwidth, Capacity, and Connectivity

Explore memory technology optimization for large-scale AI systems, focusing on bandwidth, capacity, and connectivity solutions including UCIe and photonics interconnects.

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

Open Compute Project

Interconnects and Connectivity for At-Scale AI - Shaping the Future of Interconnects and Memory

Explore cutting-edge interconnect technologies and memory solutions driving large-scale AI infrastructure development and performance optimization.

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

Open Compute Project

End to End Management of a CXL Enabled Datacenter Memory Fabric

Explore CXL technology for dynamic datacenter memory management, covering fabric architecture, composable orchestration, and real-time provisioning strategies.

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

Open Compute Project

Evaluating CXL-Based Composable Memory Solutions - A TCO Framework for Datacenter Workloads

Explore TCO frameworks for CXL-based composable memory solutions in datacenters, covering bandwidth, capacity expansion, and pooled memory efficiency for AI and analytics workloads.

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

Open Compute Project

Emerging Software Patterns for Programming New Memory Devices and Interfaces

Discover emerging software patterns for programming new memory devices and CXL interfaces, focusing on East-West communication optimization and hardware-software interface development.

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

Open Compute Project

Smashing Memory and I/O Limits - Architectures for the GenAI Era

Discover innovative system architectures that overcome memory and I/O bottlenecks in GenAI workloads through custom HBM solutions and advanced D2D-to-C2C fanout designs for scalable systems.

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

Open Compute Project

Efficient AI Serving at Scale - Processing Near Memory Acceleration for LLMs and Vector Search

Discover how processing near memory acceleration optimizes LLM and vector search performance for efficient AI serving at enterprise scale.

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

Open Compute Project

Server Scale Up Interconnect Technologies - UALink SuperNode and CXL Memory Pool

Explore datacenter scale-up interconnect technologies, UALink protocol features for GPU clusters, and CXL memory pooling solutions offering 10TB-level capacity for cloud databases.

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

Open Compute Project

Multi-Headed Composable Memory Appliance MHCMA Next Generation

Explore next-generation ASIC-based composable memory architecture with 8x8 PCIe-Gen5 CXL2.0 host ports per blade, advancing from FPGA to 12nm ASIC for enhanced pooled memory solutions.

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

Open Compute Project

Large CXL Memory Expansion Systems and Workloads

Discover how CXL memory expansion systems enable 100 TiB-class workloads through composable provisioning, shared memory paradigms, and breakthrough performance architectures.

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

Open Compute Project

Conceptualizing Next Generation Memory and Storage Optimized for AI Inference

Explore next-generation memory and storage concepts optimized for AI inference, including Processing-In-Memory (PIM) technology and novel architectures for LLMs.

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

Open Compute Project

Heterogeneous Memory System Architecture for Next-Generation Agentic AI Server Platform

Explore cutting-edge memory architecture solutions for AI servers, featuring SOCAMM, MRDIMM, and CXL technologies to address growing capacity demands and energy efficiency challenges.

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

Open Compute Project

DC Memory Fabric Management - An In-Depth Review of the Kubernetes OCP Reference Implementation

Explore how Kubernetes orchestrates CXL-attached memory through the Data Center Memory Fabric Manager, bridging white paper concepts with real-world implementation examples.

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