Scalable Parallel Computing Lab, SPCL @ ETH Zurich Courses

Productive Performance Engineering for Weather and Climate Modeling with Python

Optimize weather and climate models using Python-embedded DSL and data-centric techniques. Improve productivity and performance through abstraction, semi-automated workflows, and transfer tuning for efficient scaling on supercomputers.

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

Interactive Data Movement Visualization for Program Performance Optimization

Explore interactive data movement visualization for program performance optimization. Analyze dataflow, reuse patterns, and optimize applications without execution using SDFG IR and static analysis techniques.

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

AI Engine Architecture: Data Movement, Synchronization, Reconfiguration and Application Mapping

Explore AMD/Xilinx AI Engine architecture: vector processors, data movement, synchronization, and application mapping for machine learning and 5G acceleration.

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

Heterogeneous Serverless Computing for High-Performance Computing and AI Workloads

Explore heterogeneous serverless computing for efficient HPC, HPDA, and AI workloads. Learn about innovative accelerator uses and flexible deployment models from edge to exascale.

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

Next-Generation Networks for Machine Learning

Explore next-gen networks for machine learning, focusing on optical fabrics and scheduling techniques to accelerate distributed DNN training workloads and overcome datacenter network bottlenecks.

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

Self-Adjusting Networks for Optimized Datacenter Topology Design

Explore self-adjusting networks optimized for datacenter traffic, including metrics, design principles, and research challenges. Gain insights into demand-aware graphs and related optimization problems.

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

Post-Moore Spatial Computing: From Chips to Clusters

Explore spatial computing strategies for post-Moore's Law era, addressing challenges in scaling compute power from chips to clusters for AI and high-performance computing.

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

Portable High-Performance Python on CPUs, GPUs, and FPGAs

Optimize Python for high performance on CPUs, GPUs, and FPGAs. Learn HPC-oriented extensions, automatic optimizations, and portable performance techniques across architectures.

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

Automating Distributed Heterogeneous Computing for Domain Experts

Explore new approaches for enabling domain experts to program future HPC systems efficiently, focusing on compiler and runtime techniques for distributed heterogeneous computing.

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

Neural Graph Databases: Combining GNNs with Labeled Property Graphs

Explore neural graph databases, combining graph databases with graph neural networks using LPG2vec encoder for improved prediction accuracy in complex graph datasets.

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

LRSCwait: Enabling Scalable and Efficient Synchronization in Manycore Systems

Explore scalable synchronization in manycore systems with LRSCwait. Learn about polling-free and retry-free operations for improved throughput and energy efficiency in shared-memory architectures.

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

Compressing Multidimensional Weather and Climate Data Into Neural Networks

Compress multidimensional weather and climate data into neural networks, achieving high compression ratios while preserving accuracy for climate change analysis and severe weather prediction.

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

HexaMesh: Scaling to Hundreds of Chiplets with an Optimized Chiplet Arrangement

Explore HexaMesh, an optimized chiplet arrangement for 2.5D integration, outperforming grid layouts in network diameter, bisection bandwidth, latency, and throughput for large-scale designs.

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

Adjusting Network-on-Chip Topologies for Design Goals and Architectures

Explore a novel Network-on-Chip topology for scalable chip design. Learn design principles, cost-performance trade-offs, and toolchain development for customizing NoC architectures to meet diverse goals.

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

Performance-Detective: Automatic Deduction of Cheap and Accurate Performance Models

Discover how Performance-Detective reduces performance modeling costs while maintaining accuracy. Learn to deduce parameter interactions and derive efficient experiment designs for optimizing application configurations.

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