Future Real-Time Applications Based in Mobile Clouds and Software-Defined Networks
Explore real-time online applications with Sergei Gorlatch. Learn about challenges, innovative frameworks, cloud computing, and software-defined networking for optimizing performance and user experience.
Dreams vs. Reality of Applied Data Science in Observability Space
Explore challenges and lessons learned in applying data science to service observability, focusing on real-world project experiences and practical insights.
The Strangler Fig Pattern for Microservices Migration
Explore the Strangler Fig pattern for migrating monolithic applications to microservices. Learn practical steps, pros and cons, and tackle technical challenges in legacy modernization.
Kiali Beyond the Graph - Troubleshooting Istio
Explore Kiali's advanced features for observability and management of Kubernetes microservices, focusing on troubleshooting Istio service mesh complexities.
Running ML Workloads with AWS Purpose-Built ML Accelerators and Ray
Discover AWS Trainium and Inferentia ML accelerators for high-performance, cost-effective Generative AI in the cloud. Explore native support for these accelerators in Ray.
Llama - Scaling Up LLMs in an Open Ecosystem
Explore Llama's evolution, its impact on Generative AI, and learn how to leverage state-of-the-art performance using Ray-based platforms for building with open-source large language models.
Optimizing LLM Training with Airbnb's Next-Gen ML Platform
Explore Airbnb's ML platform upgrade for efficient LLM use, featuring Ray integration for improved developer productivity, performance, and cost-effective GPU utilization on AWS.
Managed Cloud Infrastructure for LLMs
Explore solutions for LLM infrastructure challenges, including cost reduction, GPU availability, scalability, and cloud resource management. Learn to accelerate AI development and deployment.
OneDigital ML Platform: Unleashing the Power of Machine Learning at Scale
Explore OneDigital's journey in building a scalable ML platform using Ray and Anyscale. Learn about architecture, strategies, and retail use cases for advanced ML models.
KubeRay: A Ray Cluster Management Solution on Kubernetes
Explore KubeRay's integration with Kubernetes, its capabilities for Ray cluster management, and performance benchmarks for simplified deployment of Ray applications.
Scaling AI Health Assistants: Challenges and Solutions
Explore challenges and solutions in scaling AI health assistants, focusing on model management, state persistence, and system reliability. Learn key principles for building scalable AI products using Ray.
Accelerating Data Serialization and Deserialization Protocols with In-Network Compute
Explore in-network compute techniques to accelerate data serialization/deserialization protocols, enhancing efficiency in datacenter and HPC systems. Learn about SmartNIC offloading and its impact on performance.
Swing: Short-cutting Rings for Higher Bandwidth Allreduce
Optimize allreduce performance on torus networks with Swing algorithm. Improve bandwidth by reducing distance between communicating nodes, outperforming existing methods by up to 3x.
JuiceFS CSI in Multi-Thousand Node Kubernetes Clusters for LLM Pre-Training
Explore JuiceFS CSI's innovative solutions for managing large-scale data in Kubernetes, addressing storage challenges in LLM pre-training with mount pod and sidecar modes for optimal performance.
XRegistry - Looking Beyond CloudEvents
Explore xRegistry's role in enhancing CloudEvents for distributed systems, enabling efficient resource discovery, metadata querying, and automation in event-driven architectures.
DRust - Language-Guided Distributed Shared Memory with Fine Granularity and Full Transparency
Explores DRust, a Rust-based distributed shared memory system that leverages language ownership semantics for improved efficiency and scalability in distributed computing environments.
Beaver: Practical Partial Snapshots for Distributed Cloud Services
Innovative protocol for capturing partial snapshots in distributed cloud services, ensuring causal consistency without disrupting operations. Leverages load balancers for minimal changes and near-zero overhead.
Inductive Invariants for Streamlining Distributed Protocol Proofs
Streamlining distributed protocol proofs using invariant taxonomies. Simplifies finding inductive invariants by categorizing them into Regular and Protocol Invariants, reducing developer effort in proving correctness.
Fast and Scalable In-network Lock Management Using Lock Fission
Innovative distributed lock service using programmable switches for improved latency and throughput, supporting millions of locks through lock fission and on-demand migration techniques.
Motor: Enabling Multi-Versioning for Distributed Transactions on Disaggregated Memory
Innovative system for efficient distributed transactions on disaggregated memory, improving throughput and reducing latency through multi-versioning and RDMA-based protocols.
Chop Chop: Byzantine Atomic Broadcast to the Network Limit
Innovative Byzantine Atomic Broadcast system achieving high throughput and low latency through novel batching technique. Demonstrates superior performance in geo-distributed deployments for various applications.