MoonBit, a new WebAssembly‑based programming language from China, targets cloud and edge computing with a vertically integrated toolchain, offering tiny WASM binaries, high runtime performance, advanced compilation speed, and a data‑oriented design, and includes sample code, benchmarks, and future roadmap details.
MoonBit, a new WebAssembly‑based programming language created by a Chinese development team, targets cloud and edge computing with an integrated toolchain that includes a compiler, build system, IDE and deployment tools, offering smaller binaries, higher runtime performance, and advanced compile‑time optimizations, as demonstrated by benchmarked Fibonacci implementations in MoonBit, Go, and Rust.
This article examines the performance difference between Java's List.sort() and Stream.sorted() by presenting simple demos, discussing JIT warm‑up effects, and using JMH benchmarks to show that native list sorting generally outperforms the stream‑based approach.
Meta introduced Code Llama, a Llama 2‑based AI coding model available in base, Python‑specific, and instruction‑tuned versions across 7B, 13B, and 34B sizes, claiming superior benchmark performance and free community licensing for research and commercial use.
This article explains how repeatedly creating Jackson ObjectMapper instances hurts performance, demonstrates a JMH benchmark comparing new, singleton, and ThreadLocal ObjectMapper usages, and concludes that a single globally shared ObjectMapper can achieve up to ten times faster JSON parsing.
This article explains how CPU core count, frequency, cache, GPU shader count, memory bandwidth, and graphics card VRAM affect performance, and presents 2023 desktop and laptop performance ladder charts to help users compare and choose suitable hardware for their needs.
This article investigates the performance difference between Java's native List.sort() method and the Stream.sorted() approach, provides simple demos, explains JIT warm‑up effects, and presents JMH benchmark results that show List.sort() consistently outperforms Stream.sorted() for typical collection sizes.
This article investigates whether Java's list.sort() truly outperforms stream().sorted() by presenting simple demos, explaining the pitfalls of naive timing, and using JMH benchmarks to reveal that list.sort consistently runs faster due to lower overhead, while the performance gap remains small for typical data sizes.
This comprehensive guide introduces Linux perf, explains its core components, walks through essential commands, demonstrates real‑world use cases such as CPU, memory, and I/O analysis, shows how to generate flame graphs, and provides advanced tips and troubleshooting for accurate performance profiling on Linux systems.
The 2023 Fairwinds Kubernetes benchmark, analyzing over 150,000 workloads, reveals that many organizations still miss critical best‑practice configurations such as memory limits, liveness probes, proper image pull policies, replica counts, and CPU limits or requests, leading to increased security risks, uncontrolled cloud costs, and reduced reliability.
Webman is a high‑performance PHP HTTP framework built on Workerman that replaces the php‑fpm model, offering minimal core, extreme speed, easy extensibility, and detailed stress‑testing guidance, making it several times faster than traditional PHP stacks and even comparable Go frameworks.
This article benchmarks async versus multithreaded programming across Rust, Go, Java, C#, Node.js, Python, and Elixir, measuring peak memory usage for 1, 10 000, 100 000 and 1 000 000 concurrent tasks to reveal surprising scalability trade‑offs.
Apache Doris 2.0 introduces a suite of high‑concurrency data‑serving enhancements—including row‑store format, partition‑bucket pruning, advanced indexing, materialized views, runtime filters, TOPN optimization, short‑circuit point‑query paths, prepared statements, and row cache—enabling single‑node tens of thousands QPS and dramatically reducing query latency.
This article presents a comprehensive overview of Arctic's streaming capabilities on Flink, detailing its mixed‑format architecture, core principles, benchmark comparisons with Iceberg, future development plans, and a Q&A session covering implementation nuances and performance considerations.
This article translates and expands on a HikariCP wiki post, showing how dramatically reducing database connection pool size—from thousands to just a handful—can cut response times from around 100 ms to 3 ms, backed by Oracle and PostgreSQL benchmarks and a practical sizing formula.
Dragonfly, an open‑source C/C++ memory cache system released by a former Google and Amazon engineer, claims superior speed and lower memory usage than Redis, but Redis’s team counters with benchmark results showing higher throughput and discusses architectural differences, scaling strategies, and future feature considerations.
The 2023 desktop CPU ladder chart ranks Intel and AMD processors by multi‑core performance, helping users quickly compare generations and choose the most suitable CPU for gaming or productivity based on benchmark rankings and recommended motherboard pairings.
This article benchmarks the Groovy 'as' type‑conversion keyword against native Java conversion methods for String‑to‑double, double‑to‑String, and double‑to‑int operations using JMH, revealing that Java's built‑in approaches consistently outperform Groovy's 'as' keyword across various test cases.
This article explores Redis’s evolution from its 2009 inception, outlines major version releases, examines its in‑memory design, efficient data structures, encoding schemes, single‑threaded event loop with epoll, and benchmark results, illustrating why Redis achieves exceptionally high throughput and low latency in real‑world deployments.
This article compares Java's Stream.toList(), Collectors.toList(), and Collectors.toUnmodifiableList() by explaining their differences, showing benchmark code, presenting performance results for both 1,000 and 10,000 element streams, and concluding which method offers superior speed and efficiency.
This article demonstrates how reorganizing Java matrix‑multiplication loops and understanding Java 2‑D array storage and CPU cache hierarchies can turn a naïve implementation into a version that runs up to a hundred times faster, backed by JMH benchmark results.
This article examines how strict adherence to clean‑code principles—such as using polymorphism, keeping functions tiny, and avoiding switch statements—can dramatically increase the number of CPU cycles required for simple area‑calculation loops, presenting benchmark results that show performance penalties of up to fifteen‑fold.
This article investigates the common belief that using try-catch in Java severely degrades performance by examining JVM exception handling, bytecode generation, compilation optimizations, and extensive benchmark tests, ultimately showing that the overhead is negligible in most scenarios.
A practical benchmark compares placing a try‑catch inside a Java for‑loop versus wrapping the loop with try‑catch, revealing that the performance difference is negligible and that the real impact depends on exception frequency and code design rather than the try‑catch placement.
This article explains how fastjson2 achieves significant performance gains by replacing reflection with Java 8 Lambda‑generated function mappings, applying zero‑copy String construction, and implementing hand‑crafted parsers for common types such as dates, backed by benchmark data.
The study compares JavaScript (via the JVM engine), Lua, and pure Java for secure, lightweight dynamic scripting in Java projects, finding that native Java is orders of magnitude faster, JavaScript outperforms Lua in complex loops when the engine is reused, while Lua remains suitable for simple scripts but all scripting options are significantly slower than compiled Java.
To ensure Kun’s Flutter‑based container upgrades improve Xianyu’s user experience, we introduced a layered performance testing framework that defines industry‑standard metrics, component‑ and business‑level benchmarks, multi‑dimensional stress and regression checks, automated scripts, expanded device pools, and streamlined task creation, delivering faster insights, higher stability, and actionable optimization feedback.
This article presents a detailed performance benchmark of the new open‑source memory cache Dragonfly against Redis 7.0, describing test methodology, hardware setup, command configurations, and the resulting throughput and latency differences across various GET and SET workloads.
This article explains the fundamental differences between creating objects with the new operator and using Java reflection, presents benchmark code showing a large performance gap, analyzes the reasons behind the slowdown, and outlines practical scenarios where reflection is appropriate.
The new RocksDB 7.5.3 release dramatically reduces write‑stall time, lock contention and CPU usage while improving OPS and latency compared with 6.2.9.x, as shown by detailed memtier benchmark tests on a 4‑CPU, 32 GiB VM with NVMe storage.
By examining benchmark scores, Google Trends, and the TIOBE index, this article evaluates the performance, energy consumption, and popularity of languages such as C, C++, Rust, Java, Python, and Go, ultimately arguing that no single language will reign supreme.
This article explains how ByteDance’s frontend monitoring SDK is continuously optimized for bundle size and runtime performance through micro‑ and macro‑level code refactoring, benchmark testing with Benny and Puppeteer, and real‑world analysis using Perfsee Lab, ultimately achieving minimal impact on page load metrics.
A simple benchmark using Bubble Sort on a 10,000‑element random array compares several popular programming languages, revealing that Python performs the worst while Go (Golang) delivers a striking performance advantage even over Node.js.
This article examines the open‑source memory cache Dragonfly, its performance claims against Redis, presents reproduced benchmark results, discusses architectural differences, and outlines Redis’s design principles and testing methodology.
This article explains how ByteDance's frontend monitoring SDK was systematically reduced in bundle size and runtime overhead through micro‑optimizations such as avoiding long class names, using functional patterns, minimizing polyfills, splitting files, leveraging Puppeteer‑based benchmarks, and applying requestIdleCallback scheduling to improve user experience on both web and mobile platforms.
SelectDB, a new cloud‑native data warehouse built on Apache Doris, topped the ClickBench analytical database benchmark on c6a.4xlarge 500 GB gp2 instances, outpacing ClickHouse by 35% in Hot Run and 25% in Cold Run while also delivering high write throughput of over 140 MB/s.
This article examines the open‑source memory cache Dragonfly, compares its performance and architecture with Redis through detailed benchmark reproductions, discusses the underlying design trade‑offs, and highlights Redis’s continued relevance and future directions in real‑world deployments.
This article investigates the common belief that Java's try‑catch blocks severely degrade performance by examining JVM exception handling mechanisms, analyzing compiled bytecode, and presenting extensive benchmark tests under various JVM compilation modes to determine the real impact of exception handling on execution speed.
This article examines the open‑source memory cache Dragonfly, compares its performance and architecture against Redis through detailed benchmark results, discusses Redis’s response and design principles, and provides reproducible test configurations and command lines for both systems.
A comprehensive benchmark shows RedisJSON delivering dramatically higher throughput and lower latency than MongoDB and ElasticSearch across isolated writes, isolated reads, and mixed workloads, with up to 200‑fold faster writes and 500‑plus‑fold faster reads, while maintaining sub‑millisecond response times.
An in‑depth comparison shows that while Dragonfly claims to be the world’s fastest memory cache, Redis 7.0 often delivers higher throughput and lower latency across multiple benchmark scenarios, highlighting architectural trade‑offs and the potential for future feature adoption from Dragonfly into Redis.
This article compares the open‑source memory cache Dragonfly, created by a former Google and Amazon engineer, with Redis 7.0, analyzing benchmark results, architectural differences, and community reactions to determine which system delivers superior performance and scalability.
This article introduces the multimodal Rastrigin function, explains its challenges for optimization, and demonstrates a complete Python implementation of a genetic algorithm—including encoding, selection, crossover, mutation, and decoding—to locate the function’s global minimum, with visual results and performance analysis.
This article investigates the common belief that Java try‑catch blocks severely degrade performance by examining JVM exception handling mechanics, compiled bytecode, tiered and JIT compilation, and extensive benchmark tests, ultimately showing that try‑catch has negligible impact when no exception occurs.
The article examines the open‑source Dragonfly memory cache, its performance claims versus Redis, the Redis team's rebuttal, detailed benchmark methodology, results across GET/SET workloads, and the architectural principles that keep Redis competitive in real‑time in‑memory data storage.
A comprehensive AWS benchmark reveals that RedisJSON dramatically outperforms MongoDB and ElasticSearch across isolated writes, isolated reads, mixed workloads, ops‑per‑second, and latency, maintaining sub‑millisecond response even under high update rates, while the other databases suffer significant slowdowns.
The article explains perfect hash functions, their collision‑free construction methods such as FCH, CHD, and PTHash, compares them to conventional hash tables, reviews common and cryptographic hash functions, and shows how read‑only perfect‑hash maps deliver faster lookups and lower memory use for static key sets.
This article examines the open‑source Dragonfly memory cache, compares its performance and architecture against Redis using reproducible benchmarks on AWS c6gn.16xlarge instances, and discusses the implications of the results for future in‑memory data store designs.
The article examines the open‑source release of Delta 2.0, compares its features and benchmark results with Iceberg and Hudi, discusses the core capabilities required by enterprises for a lakehouse architecture, and introduces the Arctic project as a multi‑engine streaming lake service.
The article reviews recent developments in data‑lake table formats—Delta Lake 2.0, Iceberg, and Hudi—examining their features, benchmark results, and ecosystem impact, and then introduces Arctic, an open‑source streaming lakehouse service built on Iceberg that aims to bridge batch‑stream gaps for enterprises.
The ip2region 2.10.0 update introduces C# and Node.js query clients, provides detailed benchmark results, fixes a Python XDB searcher bug, disables buffer‑related functions for several languages, and includes a Maven dependency snippet for Java, all while maintaining microsecond‑level query performance for offline IP lookup.
This article examines the implementation of java.util.concurrent.DelayQueue for high‑throughput performance testing, presents a custom Delayed object, shares benchmark code and results across various thread counts, and concludes that DelayQueue can sustain several million QPS when properly tuned.
This article examines Fastjson2's performance compared to Fastjson, detailing environment setup, benchmark results, GC analysis, and a deep dive into the library's serialization and deserialization source code to explain why memory settings dramatically affect speed.
A recent paper examines the energy consumption, runtime, and memory usage of software across 27 programming languages, revealing that compiled languages like C and C++ are the most energy‑efficient, while highlighting the growing importance of energy awareness in software development and its implications for embedded, mobile, and cryptocurrency applications.
MegPeak is an open‑source processor debugging tool that evaluates instruction peak bandwidth, latency, and memory bandwidth, enabling developers to benchmark CPUs, plot Roofline models, and guide low‑level algorithm and kernel optimizations for AI workloads.
This article presents a detailed benchmark of java.util.concurrent.ConcurrentHashMap under high thread counts, explains the test methodology using a Groovy‑based FunTester framework, shares raw performance numbers, and uncovers that CPU limits and random number generation are the primary bottlenecks.
The article presents a detailed benchmark of AMD's Milan‑X 7773X processor, highlighting its 3D V‑Cache architecture and large 768 MB L3 cache, and demonstrates up to 1.80× speed‑up for HPC applications WRF and OpenFOAM on a dual‑socket NF5468A5 GPU server.
The article reviews high‑performance networking options—RDMA (including RoCE v2 and iWARP) and Linux’s io_uring—explaining their principles, hardware requirements, and benchmark results, and concludes that while RDMA delivers ultra‑low latency for specialized workloads, io_uring offers modest network benefits, leaving TCP as the default for most services.
The LDBC has approved the world’s first financial graph database benchmark, FinBench, a collaborative effort led by Ant Group to provide a rigorous, open‑source testing suite that simulates real‑world financial workloads and fills a critical gap in graph database evaluation.
This article reviews the seminal SRCNN paper, detailing its contributions, architecture, training pipeline, hyper‑parameters, and extensive experiments that show how a shallow fully‑convolutional network achieves superior PSNR and runtime compared to traditional sparse‑coding and bicubic methods.
The Python team released a beta of Python 3.11, and benchmark tests using Docker on an M1 Pro MacBook Pro show the new version averaging 14% faster than 3.10 and achieving up to 64% speed improvements in many workloads, highlighting notable performance enhancements.
A recent benchmark using the pyperformance suite shows that Python 3.11 beta runs up to 64% faster than Python 3.10 on most tests, with an overall 14% speed increase, highlighting the language’s long‑awaited performance improvements.
A comprehensive performance benchmark shows RedisJSON delivering dramatically higher throughput and lower latency than MongoDB and ElasticSearch across isolated writes, isolated reads, and mixed workloads, highlighting its superior scalability and stability for modern data‑intensive applications.
This article examines the performance impact of instantiating a new Jackson ObjectMapper for each JSON parsing operation versus reusing a global instance, using JMH micro‑benchmarks to show that a shared ObjectMapper can be up to ten times faster.
This article benchmarks three ObjectMapper usage patterns—creating a new instance per call, sharing a global singleton, and using ThreadLocal—demonstrating that a single shared ObjectMapper can achieve up to ten‑fold higher throughput in JSON parsing.
This article, the second in a Go high‑performance series, details concurrency optimizations including lock‑free data structures versus locked lists, sharding and RWMutex to cut lock contention, controlling goroutine creation with pooling, using sync.Once for cheap one‑time initialization, and employing sync.Cond for efficient goroutine notification.
The author recounts submitting a Go PR to fix a benchmark timeout overflow bug, detailing the code review feedback, overflow detection issues, testing strategies, and proper commit‑message conventions, ultimately sharing insights from a largely unsuccessful contribution attempt.
Community testing of StarRocks 2.0 revealed that, across multiple benchmarks—including low‑cardinality queries, SSB workloads, and high‑concurrency scenarios—StarRocks consistently outperformed ClickHouse and Druid, delivering performance gains ranging from 2‑3× to over 7×.
The author recounts a failed Go pull‑request, detailing overflow bugs, missing tests, improper commit messages, and how proper code review practices can turn a buggy contribution into a successful merge.
A comprehensive Phoronix benchmark on a Raspberry Pi 400 reveals that the new 64‑bit Raspberry Pi OS consistently outperforms its 32‑bit counterpart, delivering performance gains ranging from modest single‑digit improvements to massive jumps of over 1300% depending on the workload.
This article presents a detailed benchmark comparing MySQL 8.0.19 and MariaDB 10.4.12 on a Windows 10 i7 machine, covering single‑row and batch inserts, various SELECT queries with and without indexes, and analyzes the trade‑off between speed and memory usage.
This article presents a comprehensive performance comparison between LMAX Disruptor and Java's LinkedBlockingQueue, detailing test setups, producer and consumer configurations, various object sizes, benchmark results, and practical conclusions for high‑throughput backend systems.
This article examines various methods for extracting subarrays in Swift, benchmarking loop‑based copying, range subscripting with map, while loops, and direct memory copying via memcpy, analyzing their execution times, compiler optimizations, and trade‑offs to guide developers toward efficient array handling in performance‑critical applications.
This article compares Python's while and for loops, shows benchmark results revealing that for loops run faster due to fewer Python‑level operations, and demonstrates that using built‑in functions like sum or applying a mathematical formula can make looping dramatically faster, often eliminating the loop entirely.
Alibaba Cloud’s self‑developed MRACC (Apasara Compute MapReduce Accelerator) leveraged hardware‑software integration, Spark and Hadoop optimizations, and eRDMA networking to achieve the top TPCx‑BB SF3000 performance, delivering up to 2‑3× faster SQL queries and 30% faster Spark shuffle, with significant cost efficiency gains.
Researchers at ETH Zurich introduce Exemplar Transformers, a novel Transformer layer that accelerates visual object tracking by eight times, runs in real‑time on CPUs, and improves robustness when integrated into a Siamese‑based tracker, achieving state‑of‑the‑art performance on six benchmark datasets.
This article explains how to use Go's test command with various flags such as -c, -run, -timeout, and -coverprofile, demonstrates writing test cases, TestMain setup, benchmarking, profiling with pprof, generating flame graphs, and interpreting coverage reports to improve Go application reliability and performance.
The report presents a comprehensive performance benchmark of RedisJSON (RediSearch) against MongoDB and ElasticSearch, showing RedisJSON’s superior write, read, and mixed‑workload throughput and latency across isolated and hybrid scenarios, with detailed test methodology and analysis.
Recent benchmark tests show RedisJSON (with RediSearch) dramatically outperforms MongoDB and ElasticSearch, delivering up to 5.4‑times faster isolated writes, over 200‑times faster than ElasticSearch, 12.7‑times faster reads, and significantly lower latency and higher throughput across mixed workloads.
The report presents extensive YCSB‑based performance benchmarks showing that RedisJSON (with RediSearch) outperforms MongoDB and ElasticSearch by large margins in isolated writes, isolated reads, mixed workloads, and latency across a variety of operation mixes on identical AWS hardware.
This article presents a detailed performance benchmark of RedisJSON (RediSearch) against MongoDB and ElasticSearch, showing RedisJSON’s superior throughput and latency across isolated writes, reads, and mixed workloads, supported by YCSB test results and extensive latency analysis.
This article walks through using the FunTester framework to benchmark MySQL insert and delete operations, showing Java code examples, test configurations, and JSON results that reveal a high QPS for inserts and a noticeable slowdown when deletions are performed.
A recent analysis comparing 27 programming languages shows that despite Python's popularity, it and Perl consume far more energy than languages like C, C++, and Rust, challenging common assumptions about performance and environmental impact.
The article announces the official release of PHP 8.1 and presents a detailed performance benchmark conducted on an Ubuntu Linux AMD EPYC 72F3 server, comparing multiple PHP versions and highlighting the roughly 4% speed improvement of PHP 8.1 over its predecessor.
An extensive Phoronix benchmark on Ubuntu‑based AMD EPYC servers compares PHP 8.1 with earlier releases, revealing roughly 4 % speed gains over PHP 8.0, lower peak memory usage, and consistent performance improvements across micro‑benchmarks, XML parsing, and JSON generation.
The article presents a detailed performance comparison of three HTTP client implementations—Java FunTester, Go's net/http, and the high‑performance fasthttp library—across multiple thread counts and artificial latency settings, reporting CPU usage, memory consumption, and QPS to determine the most efficient choice.
This article introduces Java 8's Stream API, explains its intermediate and terminal operations, highlights its functional characteristics, compares its performance with traditional iterator loops through benchmark tests, and offers practical recommendations for using streams and parallel streams in backend development.
This guide explains how to use Redis's built-in redis-benchmark tool to simulate concurrent client load, interpret key performance metrics such as request latency and throughput, and monitor server resource usage, helping operators prevent cache-related failures like penetration and avalanche after deployment.
The article compares Go's built‑in encoding/json with third‑party libraries jsoniter, easyjson, jsonparser, and the author's jsonvalue, outlining each one's features, performance benchmarks, and ideal use cases—standard library for simplicity, easyjson for maximum speed with code generation, jsonparser or jsoniter for selective extraction, and jsonvalue for flexible, case‑insensitive handling.
This article examines Dubbo’s connection‑control feature, demonstrates how long‑living TCP connections are created, runs benchmark tests with one and two connections, discovers that varying IO thread counts skew results, and shows that increasing connections alone rarely improves throughput.
This article walks through a practical performance test for continuously adding key‑value pairs to Redis using the FunTester framework, detailing the test design, Java and Groovy implementations, execution logs, result metrics, and key takeaways for future Redis benchmarking.
This article provides a comprehensive tutorial on setting up a one‑master‑one‑slave Redis replication environment, configuring read‑write separation, verifying data synchronization, testing read‑only behavior, and performing performance benchmarking with redis‑benchmark, complete with command examples and result analysis.
This article analyzes Kuaishou's short‑video quality advantages by detailing its production‑consumption chain, multi‑dimensional evaluation methodology, objective and subjective test results, and key metrics such as bitrate savings, VMAF gains, win and selection rates, supported by visual examples.
This article explains how to use Python's multiprocessing module to parallelize a custom preprocessing function on a 537k‑record dataset, demonstrating roughly a thirty‑fold reduction in execution time compared with single‑process execution.
The article examines the rapid growth of recommendation systems, highlighting the need for industrial‑grade benchmarks, transparent explainability, and addressing algorithmic confounding caused by feedback loops, while discussing how these issues affect both users and content providers in the AI‑driven ecosystem.
This article systematically presents the cold‑start performance problems of the Beike iOS app, defines testing standards, explains the essence of optimization, lists common pitfalls, and details a comprehensive set of practical solutions ranging from lifecycle‑aware task scheduling and minimal‑set launchers to dynamic‑library lazy loading, compile‑time I/O elimination, static‑initializer handling, dead‑code removal, and monitoring standards.
The article describes how Beike Real Estate consolidated multiple OLAP engines into a single DorisDB‑based platform, detailing the business challenges, DorisDB’s technical advantages, extensive performance and concurrency benchmarks, and the resulting improvements in stability, query speed, and operational simplicity across various business scenarios.
This article introduces the CBLUE benchmark, outlines its eight Chinese medical NLP tasks, reviews the datasets and baseline Chinese pretrained language models, and analyzes performance results to highlight challenges and future directions for AI in healthcare.
This article explains Java 8 Stream fundamentals, compares its intermediate and terminal operations with traditional iterator loops through a series of benchmark tests (mapping, filtering, sorting, reduction, string joining, and mixed operations), analyzes results for different data sizes and CPU configurations, and provides practical guidance on when to use Stream, parallel Stream, or iterator in backend development.
This article explains how Hopsworks’ feature store unifies online low‑latency and offline high‑throughput storage using a dual‑system architecture built on RonDB, detailing its API, metadata handling, ingestion pipeline, benchmarks, and how it simplifies production machine‑learning feature access.
