NHR@FAU Courses

Slurm Batch Scheduler Best Practices and Advanced Use

1 rating

Master advanced Slurm batch scheduling techniques, from multi-GPU jobs and dependencies to data staging and workflow optimization for high-performance computing clusters.

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

From Picojoules to Gigawatt-hours - Energy-to-Completion and GPU DVFS for LLM Workloads

Explore energy optimization in LLM workloads from picojoules to gigawatt-hours, focusing on GPU DVFS techniques and energy-to-completion metrics for transformer models.

• YouTube
• 1 hour 5 minutes
• Self-Paced
• Free Video

Inference in the Age of Reasoning Models

Explore distributed inference techniques for large language models, focusing on architectural optimizations like KV caching and prefix reuse to enable scalable reasoning AI deployment.

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

Parallel Programming 2020

Master OpenMP and MPI programming techniques for high-performance computing through comprehensive lectures covering computer architecture, parallelism concepts, and hybrid programming approaches.

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

Domain Decomposition-Based Preconditioners and Neural Networks

Explore domain decomposition methods bridging numerical simulation and machine learning for scalable high-performance computing solutions.

• YouTube
• 1 hour
• Self-Paced
• Free Video

Compiler Architecture - Abstractions for Domains, Accelerators, and Microarchitecture

Explore compiler design principles, domain-specific languages, accelerator targeting, and optimization techniques for high-performance computing applications.

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

Energy Monitoring and Optimization Made Easy and Accessible - Powered by EAR

Discover EAR's vendor-independent energy monitoring system that simplifies optimization across any architecture, making advanced energy management accessible for data centers.

• YouTube
• 1 hour 4 minutes
• Self-Paced
• Free Video

Assessing the Efficiency of Your Jobs with ClusterCockpit

Master ClusterCockpit to monitor HPC job performance in real-time, analyzing load balance, flop/s rates, communication patterns, and I/O efficiency for optimal resource utilization.

• YouTube
• 1 hour 5 minutes
• Self-Paced
• Free Video

Memories of a Performance Engineer - 38 Years of Tools Building

Discover the evolution of HPC performance tools through 38 years of development, from TDL/POET to Score-P, with insights on successes, failures, and the current state of parallel systems analysis.

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

Code Generation Technology for HPC Codes - Case Studies on waLBerla and HyTeG

Explore code generation technology for HPC applications through waLBerla and HyTeG case studies, focusing on performance portability and optimization strategies.

• YouTube
• 1 hour 4 minutes
• Self-Paced
• Free Video

Scaling and Accelerating LLM Trainings

Explore scaling laws, parallelization techniques, and low-precision training methods to maximize efficiency in large language model training on high-performance clusters.

• YouTube
• 1 hour 1 minute
• Self-Paced
• Free Video

Generative AI and Probabilistic Physics Simulations

Explore how generative AI methods like diffusion models enhance physics simulations by learning probability distributions and capturing uncertainty in complex systems.

• YouTube
• 1 hour 3 minutes
• Self-Paced
• Free Video

Increase Productivity with Neovim, Language Server Protocols, and Modern Terminal Tools

Discover modern terminal tools and Neovim setup to boost HPC development productivity with Language Server Protocols and best practices for efficient workflows.

• YouTube
• 1 hour
• Self-Paced
• Free Video

Quantum Computing in the Munich Quantum Valley

Explore quantum computing developments in Munich Quantum Valley, covering multi-modal systems, software integration, and HPC quantum accelerators for broad applications.

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

From Data Traces to System Insight - Redefining I/O Understanding in HPC

Explore how I/O behavior transforms from performance constraint to system insight source in modern HPC environments and data-intensive workloads.

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