Everything tagged performance-optimization, newest first. Tags come from the classifier reading each item's summary; 394 tags used 25 times or more have their own page.
Cloudflare prototyped cache compression to determine whether the same hardware could store more cached content, exploring if compressing objects inside the cache would increase effective capacity without adding infrastructure.
Five Rust-level memory optimizations to Big Pineapple’s DNS cache layout reduced per-entry memory usage by 56%, freeing roughly 100 TB across Cloudflare’s fleet.
Geoff Huston explores the unpredictable nature of DNS queries following a 'cold start', where no cached data is available, highlighting the complexities and inefficiencies that arise in such scenarios.
Context engineering, unlike prompt engineering, focuses on structuring and optimizing the information provided to AI models to enhance their performance by delivering relevant and concise data, avoiding cognitive overload and context rot.
Cache Response Rules provide a solution for managing cache behavior by addressing issues caused by headers like Set-Cookie or Cache-Control that are difficult to modify at the origin.
Meta addressed potential latency issues in its ad serving fleet caused by a Linux kernel upgrade by implementing a customized scheduling policy using the BPF-based sched_ext framework to optimize ad delivery performance.
Workers Cache is a new, regionally tiered caching solution that enhances Worker entrypoints by being infinitely composable and configurable through standard HTTP headers.
Apple's operating systems have been updated with Liquid Glass enhancements and various performance optimizations to improve user experience.
Meta's Capacity Efficiency Program utilizes AI agent platforms to automate performance issue resolution, enhancing infrastructure efficiency and enabling engineers to focus on innovation.
Enthropic's latest update introduces Claude managed agents, a comprehensive AI agent builder platform that simplifies deploying production-ready AI agents for long-running tasks, enhancing automation and efficiency with built-in intelligence and performance optimizations.
Netflix developed an experimental caching layer to manage the high query load on Apache Druid by caching older data and only querying new data, significantly reducing the load and improving performance during high-demand events.
Investigating a 30-minute restart delay in Atlantis revealed a Kubernetes volume permissions bottleneck, resolved by adjusting fsGroupChangePolicy, reducing restart times to 30 seconds.
The expansion of Regional Services introduces new pre-defined and Custom Regions, allowing customers to define specific geographical boundaries for data processing to meet compliance and performance requirements.
Netflix optimized its Ranker service by re-architecting the video serendipity scoring feature, reducing CPU usage and latency through batching, memory layout improvements, and leveraging the JDK Vector API for SIMD operations.
Netflix encountered CPU architecture bottlenecks while modernizing their container runtime, leading to significant lock contention during container scaling, prompting both hardware and software optimizations to enhance performance and reliability.
The course offers practical techniques for optimizing React app performance by understanding rerendering and employing strategies like memorization, derived state, and component splitting to enhance efficiency in production environments.
This guest post provides a detailed checklist of essential considerations for designing a network edge, serving as a foundational guide for network architects.
This article, part of a three-part series, explores the differences between rate-based and credit-based flow control methods in high-performance network design.
Netflix's Maestro workflow engine was redesigned to enhance performance by 100X, reducing overhead from seconds to milliseconds, while maintaining scalability and reliability for diverse data and ML workflows.
Disabling JavaScript optimization in browsers like Chrome, Edge, and Firefox can enhance protection against zero-day vulnerabilities, although it may slightly impact performance without noticeable effects for most users.
The privacy proxy service experienced a notable speed improvement by eliminating a 40ms delay in the "double-spend" verification process.
Organizations often store excessive data, leading to increased costs, reduced performance, and hindered sustainability, despite the seemingly safe approach of retaining data "just in case."
The week ending June 13, 2025, featured significant self-hosted news, updates, and launches, highlighting advancements and innovations in the self-hosting community.
Explore the top three home lab network topologies for beginners, featuring setup tips, free tools, and strategies to enhance performance and security.
Developers' intuitive grasp of performance becomes insufficient when dealing with the massive scale of systems like Cloudflare's.
CFKA achieves high performance by focusing on efficiency through the zero copy principle, reducing data transfer steps from disk to network, and utilizing modern hardware optimizations like DMA.
Understanding how to clean up Windows 11 is essential for optimizing its performance and maintaining a clutter-free system.
Jaime Torrealba discusses with Ryan the evolution of 3D web technologies, performance optimization, frameworks, abstraction layers, and future prospects of 3D applications in web development.
The new fork of Bolt.new, led by Cole Medan, introduces significant updates, including folder imports, mobile compatibility, and performance optimizations, enhancing its capabilities as an AI-powered web development tool.
Scalability in system design involves efficiently handling increased loads by adding resources, focusing on optimizing performance, and addressing bottlenecks through strategies like statelessness, loose coupling, and asynchronous processing.
Cloudflare transitioned from SysLog-NG to OpenTelemetry to enhance scalability, memory safety, and maintainability while ensuring uninterrupted customer traffic.
The post details performance optimizations for WAF ML, including code examples, benchmarks, and significant latency reductions after implementation.
The discussion explores CPU pipelining and introduces branch prediction, where a new robot predicts instruction addresses to optimize processing, minimizing delays when branching occurs in the instruction sequence.
