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.
Deinsum: Practically I/O Optimal Multi-Linear Algebra for Distributed Systems
Optimize distributed multilinear algebra computations with Deinsum framework. Automate data movement-optimal tiling and scheduling for enhanced performance in high-dimensional tensor operations.
Magic Castle - Terraforming the Cloud to Teach High-Performance Computing
Replicate HPC infrastructure experience using cloud resources with Magic Castle. Deploy complete clusters with schedulers, data-transfer nodes, and pre-compiled software in minutes.
Imbue - Training a 70B Model from Scratch - Infrastructure and Challenges
Explore Imbue's journey in building infrastructure for a 70B model, covering technical challenges, cluster setup, and insights into large-scale AI training with Bowei, head of infrastructure.
Open Source Industrial Control: Turning 2,800 Tons of Metal with Python and Flask
Discover how Python and Flask revolutionize hydropower generator maintenance, enhancing safety and efficiency in turning massive rotors with precision for the Bureau of Reclamation's operations.
Proxmox VE SDN VXLAN Setup Tutorial
Learn to set up Software Defined Networking with VXLAN in Proxmox VE, enabling virtual networks across cluster nodes without physical network changes. Ideal for IT professionals seeking efficient network management.
Consensus Protocols - Crypto 2024
Explore cutting-edge consensus protocols in cryptography, featuring expert presentations and discussions on the latest advancements in distributed systems and blockchain technology.
Spark RAPIDS ML - GPU Accelerated Distributed Machine Learning in Spark Clusters
GPU-accelerated machine learning in Spark using RAPIDS ML, offering significant speedup and cost savings over traditional MLlib implementations for large-scale distributed ML applications.
AT&T's Migration of Billions of Events Processing from Hadoop to Databricks
Explore AT&T's strategic shift from Hadoop to Databricks, resulting in 30% cost reduction and halved execution time for network data processing, showcasing cloud-native efficiency.
Descriptive Complexity for Distributed Computing and Neural Networks
Explore logics capturing distributed computing and neural network models, including graph neural networks. Dive into game-theoretic semantics and computational logic connections.
Pandas and Dask DataFrame 2.0 - Comparison to Spark, DuckDB and Polars
Explore Dask DataFrame 2.0's improvements, comparing its performance to Spark, DuckDB, and Polars using TPC-H benchmarks. Learn about new features enhancing speed, efficiency, and usability in big data processing.
Open Source Industrial Control: Turning 2,800 Tons of Metal with Python and Flask
Discover how Python and Flask revolutionize hydropower generator maintenance, enhancing safety and efficiency in turning massive rotors with precision for the Bureau of Reclamation's operations.
You Keep Using That Word - Asynchronous and Interprocess Comms
Explore the nuances of asynchronous communication in distributed systems, debunking misconceptions and clarifying its various interpretations to enhance understanding of inter-process communication methods.
Multi-party Computation 2 - Crypto 2024
Explore advanced multi-party computation techniques and their applications in cryptography, presented by leading researchers at a major cryptography conference.
Multi-party Computation - Crypto 2024
Explore cutting-edge developments in multi-party computation at this Crypto 2024 session, featuring presentations on advanced cryptographic techniques and protocols.
Mastering Software Operators for Data and AI - PostgreSQL, Apache Spark, and Kubeflow
Explore software operators for PostgreSQL, Apache Spark, and Kubeflow. Learn to simplify complex operations, streamline data processing, and navigate MLOps setups for optimal AI workload performance.
Federated User Representation Learning: Privacy-Preserving Personalization in Distributed Systems
Explore federated user representation learning for privacy-preserving collaborative personalization in machine learning, improving model quality while maintaining user privacy.
The Future of Large Language Models is Federated - Seminar 120
Explore cutting-edge developments in federated learning and its transformative role in the future of Large Language Models through insights from Cambridge researcher Nic Lane.
Federated Learning for Medical Image Analysis - Seminar 89
Explore cutting-edge federated learning applications in medical imaging with NVIDIA expert Holger Roth, focusing on collaborative AI development while maintaining data privacy and security.
Scaling Up Machine Learning Without Tears - Programming Languages' Role
Explore techniques for scaling machine learning workloads across accelerators, focusing on partitioning, parallelism, and programming language concepts to optimize performance and hardware utilization.
Ресурсний центр для грід- та хмарних технологій
Дізнайтеся про сучасні грід- та хмарні технології, їх застосування в науковій роботі та особливості функціонування ресурсного центру обміну даними в ІТФ НАН України.