Everything tagged performance, newest first. Tags come from the classifier reading each item's summary; 394 tags used 25 times or more have their own page.
Knight Capital’s 2012 collapse shows how a reused feature flag, inconsistent server deployment, and a mistaken rollback turned one dormant trading function into a catastrophic software bug that executed millions of bad trades in 45 minutes, costing the firm $440 million and leading to its sale.
The speaker explains how computers estimate time, from old systems that counted screen interrupts to modern CPUs with high-frequency counters, and how operating systems combine hardware clocks, boot-time offsets, and calibration to turn raw ticks into usable wall-clock time and precise timing measurements.
The content explains that strong leaderboard scores do not guarantee real-world success, because AI applications must balance accuracy, speed, and cost, and should be assessed through both model evaluations and system evaluations before users experience poor latency, errors, or scaling issues.
Apple introduced new processors together with refreshed Mac Mini and Mac Studio models, signaling an update to its desktop lineup with improved hardware options.
LINX and JPIX have expanded their 2018 partnership by launching direct London-to-Tokyo network routes, enabling European networks to access JPIX’s Japanese interconnection platform and local peering opportunities while extending cross-border connectivity between the two exchange ecosystems.
The Cloudflare Blog migration to EmDash was used to validate the stack at massive scale by stress-testing performance, safely routing live production traffic, and redesigning the frontend experience to improve reliability and usability.
Cloudflare has achieved FedRAMP Class D (High) Certified status and is pursuing DoD IL4 authorization, enhancing security and performance for the public sector.
The article explores whether various devices and systems can effectively run Jellyfin, an open-source media server, considering factors like hardware compatibility, performance, and user needs.
AI Assistant for Networking represents a new dimension in network interaction, offering a chatbot-based tool for real-time troubleshooting, configuration, and performance insights, enhancing traditional methods like CLI, GUIs, and APIs.
Microsoft claims its tools are more cost-effective and deliver better performance compared to competitors' offerings.
The content references a catchy phrase or lyrics that suggest a sequence of actions or dance moves, possibly from a song.
Moonshot AI's Kimi K3, a 2.8 trillion parameter open-source model, excels in multimodal tasks, coding, and browser use, offering top-tier performance at competitive pricing, rivaling leading AI models like GPT 5.6 Soul and Claude Fable 5.
Data center racks, traditionally chosen for dimensions and cost, now play a crucial role in performance and resilience due to increasing power, cooling, and cabling demands.
Bun, acquired by Anthropic, successfully refactored its codebase from Zigg to Rust using AI, fixing bugs, improving performance, and resolving memory issues, despite criticism from Zigg's creator.
Grock 4.5 offers frontier-level intelligence at a low cost, excelling in benchmarks like terminal bench and coding tasks, while outperforming competitors like Opus 4.8 Max and providing high-speed performance.
Guillermo Rauch emphasizes the importance of balancing price and performance when optimizing for production, as reported by TechCrunch.
In Formula One, McLaren focuses on rapid innovation and collaboration, leveraging AI and Cisco technology to enhance car performance and maintain competitive advantage under new regulations.
The introduction of a 7-angstrom transistor marks a significant advancement in semiconductor technology, offering 50% better performance or 70% power savings, and represents a breakthrough in overcoming the limitations of traditional two-dimensional scaling by exploring vertical stacking in the Z direction.
The Intel Mac era, marked by significant technological advancements and challenges, is concluding as Apple transitions to its own silicon chips.
As AI evolves, understanding its performance and costs becomes crucial, with AI observability being essential for comprehensive insights beyond mere expenses.
The upcoming release is expected to significantly enhance the performance and features of older iPhone models.
The RX 9070, a high-performance product, was launched over a year ago with a price tag of $549.
The design of a resolver's query strategy must balance the tension between resilience and speed, particularly evident during DNS cold starts.
Microsoft's latest high-end mobile workstation is a more conventional offering compared to its previous, more experimental designs.
An overview of top alternative browsers challenging Chrome and Safari highlights their unique features and potential advantages for users.
The Framework 12 offers upgradeability and a customizable port selection but falls short in performance, build quality, and display compared to the more efficient and polished MacBook Neo, despite being more expensive.
While sub-$600 laptops have been available for a long time, finding consistently high-quality options in this price range is still uncommon.
Long context and cache augmented generation (CAG) offer alternative methods to retrieval augmented generation (RAG) for providing large language models (LLMs) with external knowledge by utilizing expanded context windows and caching strategies, though each has its own advantages and challenges related to cost, latency, and performance.
Upgrading an old PC with a high-end GPU can enhance its performance and compatibility with modern gaming and applications.
Google's Gemini 3.5 Flash, launched at the Google IO conference, is a fast, efficient, and strong coding model with impressive performance but high token costs, raising questions about its efficiency as a flash tier model.
FSR 4.1, when utilized on RDNA3 or RDNA2 GPUs, could potentially lead to a more significant performance decrease.
A partitioning change in a petabyte-scale ClickHouse cluster led to stalled billing jobs due to severe lock contention in the query planner, which was resolved by identifying the issue and creating upstream patches.
The Browser Run product has been enhanced with higher usage limits, faster performance, improved reliability, and quicker shipping by leveraging Cloudflare's Containers.
Google is rapidly testing and upgrading its Gemini AI models, with significant performance improvements seen in the Gemini 3 Flash model, ahead of the Google IO conference, while also preparing broader rollouts for enterprise users.
Deep Seek version 4 introduces two open-source models, Pro and Flash, with significant parameter sizes and claims of superior performance, but real-world testing reveals limitations in creativity and execution compared to other models.
The video explores the differences between 32-bit and 64-bit applications, emphasizing that while 64-bit apps generally perform better due to greater RAM access, 32-bit versions can outperform in certain scenarios, especially when using .NET Framework apps, highlighting the importance of context in choosing between them.
The $899 chip offers limited practical benefits, making it a less compelling choice for most consumers.
Dave shares insights on the original Task Manager's development, emphasizing efficiency, minimalism, and reliability, influenced by his early programming experiences on a 1 MHz Commodore 64.
The chip's CPU dies are designed with an additional 64MB of cache stacked beneath each, enhancing performance.
The product offers excellent performance for its cost, excluding the additional expenses for RAM, SSDs, and GPUs.
The potential of eBPF to enhance performance in web servers and microservices is being explored, similar to its success in network functions.
Dynamic Workers provide a secure, lightweight environment for executing AI-generated code with millisecond startup times, significantly outperforming traditional containers.
The latest CPUs offer marginally improved speed and reduced cost compared to their predecessors.
This $599 laptop offers a budget-friendly option with some compromises, notably its 8GB RAM, but successfully balances cost and performance by cutting corners wisely.
The M5 Pro Max's "performance" CPU cores are distinct and not merely rebranded E-cores, indicating a unique design focus.
Google's Gemini 3.1 Flashlight model, designed for high-volume developer workloads, offers remarkable speed and scalability, outperforming previous models with competitive pricing and enhanced features like adjustable reasoning depth.
Bruce Davie's article critiques the push for faster-than-Dijkstra algorithms in link-state routing, highlighting that Dijkstra's algorithm efficiently handles large networks, as demonstrated by its performance on a 2000-node network in 2003.
The most notable aspect of the otherwise routine update is the surprising enhancement of the RAM capacity.
An engineer leveraged AI to recreate Next.js on Vite in a week, resulting in vinext, which builds faster, produces smaller bundles, and simplifies deployment.
The article explores whether upgrading RAM is essential for enhancing gaming performance, addressing common misconceptions about memory's role in gaming setups.
