Wie schnell kann #Java 1 Mrd. Zeilen (13,8GB) verarbeiten – ohne „Wettbewerbs-Code“? @ReneSchwietzke zeigt die ersten 80%: weniger Allokationen, bessere I/O, schnelleres Parsing.

Lies es für Praxis-Optimierungen statt Hype: https://javapro.io/de/1brc-die-ersten-80-performance/

#JVM #Performance #1BRC

Discover how #DataFrames serve as a vital tool for data-oriented programming in the #Java ecosystem.

By analyzing The One Billion Row Challenge, Vladimir Zakharov shares how Java frameworks can outperform Python in memory management while maintaining code readability.

Discover how DataFrames elevate your Java programming experience in this #InfoQ video.

🎬 Watch now | 📄 #transcript included ⇨ https://bit.ly/3MeWqDr

#DataOrientedProgramming #1BRC

Wie schnell kann #Java 1 Mrd. Zeilen (13,8GB) verarbeiten – ohne „Wettbewerbs-Code“? @ReneSchwietzke zeigt die ersten 80%: weniger Allokationen, bessere I/O, schnelleres Parsing.

Lies es für Praxis-Optimierungen statt Hype: https://javapro.io/de/1brc-die-ersten-80-performance/

#JVM #Performance #1BRC

Do you remember the #1BRC? When I asked AI back in 2024 for a solution, I got a normal stream suggestion or something similar as a plain Java loop. When I ask AI now, I get the winning solution or parts of it suggested. Even if I try to exclude these, it always wants to come back to some of these rather cutting-edge-hacky solutions.

So, no open source means no AI programming helpers or agents. More AI programmed content means no new ideas and no innovation. At least for now!

Did #AI write better Java? The experiment's back—#1BRC tests #IDE battles & hallucination. #Zencoder #JetBrains shine, but the flaws are surprising. All about context, integration #parsing. Read Steve Poole´s article now: https://javapro.io/2025/01/15/maximize-productivity-with-ai-tools-in-java-development/

#DevTools #LLM

1BRC with AI - curious and curiouser said the LLM - JAVAPRO International

Lessons from using AI tools in Java development. How to maximize productivity with AI tools through intelligent parsing and seamless IDE integration.

JAVAPRO International

Somebody did a pretty decent run at 1-billion-row-challenge with node. The author got it down to ~11 seconds (on their machine). Considering how fancy-pants rust etc people are doing it in perhaps 3-4 seconds, I'd say pretty decent.

https://jackyef.com/posts/1brc-nodejs-learnings

#nodejs #javascript #1brc

1-billion row challenge with Node.js

5 minutes to 11 seconds with Node.js is possible?

Jacky Efendi

Dive into the latest InfoQ #Java #TrendsReport with Ixchel Ruiz & Gunnar Morling!

🔍 Topics covered:
➡️ Benefits of Java's six-month release cadence
➡️ Project Lilliput & compact object headers
➡️ Nullability in Java
➡️ Python's impact on the Java ecosystem
➡️ The One Billion Row Challenge

🎧 Listen now 👉 https://bit.ly/3VP0xaF

💪 Knowledge is power! #StayAhead #StayInformed #SoftwareTrends

#InfoQ #ProjectLiliput #Python #OneBillionRowChallenge #1BRC

InfoQ Java Trends Report 2024 - Discussing Insights with Ixchel Ruiz and Gunnar Morling

In this episode, Ixchel Ruiz and Gunnar Morling sat down with podcast host Michael Redlich, Lead Editor of the Java topic at InfoQ, to discuss the recent publication of the InfoQ Java Trends Report.

InfoQ

1 billion lines of code? No problem for AI, right? An AI didn't want to program! Steve Poole convinced it. Read his surprising results: https://javapro.io/2024/11/07/1-billion-record-challenge-ai-style/

#AI #1BRC #Java #ChatGpt #ClaudeAI #Codeium #Jetbrains #JAVAPRO

1 Billion RECORD CHallenge - AI Style - JAVAPRO International

Steve Poole tested AI coding tools for bug-finding, Javadoc generation, and JUnit creation using a 1,000-line Java program. Most tools excelled in documentation but struggled with bugs and tests, except for JetBrains AI Assistant, which produced compilable code. Poole then tackled the "1 Billion Records Challenge" to retrieve and analyze massive temperature data files, iterating prompts to improve results. Codeium, Amazon Q Developer, and Microsoft CoPilot produced functional code, but many failed on large datasets. Poole found AIs useful for code review and snippets, but complex coding tasks require iterative prompting and human intervention, limiting AI's current capabilities.

JAVAPRO International
1BRC–Nerd Sniping the Java Community

Gunnar Morling discusses some of the tricks employed by the fastest solutions for processing a 13 GB input file within less than two seconds through parallelization and efficient memory access.

InfoQ

The #OneBillionRowChallenge (#1BRC) went viral in the Java community earlier this year.

In this #InfoQ talk, Gunnar Morling dives into some of the tricks employed by the fastest solutions for processing the challenge’s 13 GB input file within less than two seconds.

Expect insights into:
• Parallelization and efficient memory access
• Optimized parsing routines with SIMD/SWAR
• Custom map implementations

He also shares personal stories and key takeaways from leading this challenge for and with the community.

A must-watch video: https://bit.ly/3Yl7y3v

#transcript included

#Java #EventStreming #ChangeDataCapture #SoftwareArchitecture #DataEngineering

1BRC–Nerd Sniping the Java Community

Gunnar Morling discusses some of the tricks employed by the fastest solutions for processing a 13 GB input file within less than two seconds through parallelization and efficient memory access.

InfoQ