Apache Arrow

📣 Released today: Version 24 of the ADBC libraries. Highlights include major updates to client libraries across several languages, improvements to the Flight SQL, PostgreSQL, and SQLite drivers, and a complete overhaul of the ADBC documentation site. The new docs make it easier to find, install, and use drivers while emphasizing ADBC’s cross-language shared-library driver model. Read more in the blog post below 👇

Apache Arrow 25.0.0 is out 🎉 A few highlights from this release: • ARM64 platforms now dynamically dispatch to SVE optimized functions • ListView data can now be read from and written to Parquet files • CSV reader can skip type inference entirely with a default column type • Feather readers/writer deprecated in favor of the Arrow IPC API • New compute functions and lots of fixes and improvements 222 issues resolved across 268 commits from 66 contributors; thanks to everyone who made this release happen. And the rest of the ecosystem keeps moving: Apache Arrow JS 21.2.0 landed today, arrow-rs and arrow-go releases are in the oven. Different repos, different release schedules, but a common format! Read more: https://lnkd.in/dfTfHV7d #ApacheArrow #OpenSource #DataEngineering #Parquet

🇫🇷 Rappel : Rencontre Apache Arrow & Parquet à Paris dans 10 jours ! 🇬🇧 Reminder: Apache Arrow & Parquet meetup in Paris in 10 days! Some seats are still available for the joint Arrow & Parquet meetup we're organizing in Paris on Thursday, June 18. The meetup will be hosted in the Datadog offices in the centre of Paris. The meetup will feature the following 5 talks, as well as some time for informal discussions and networking: 📌 Zero-Materialization Merging: Concatenating Parquet Files via REE-Encoded Arrow 📌 The Sparrow ecosystem: Arrow in modern minimal C++20 📌 GDAL: integrating columnar formats into a row-oriented framework 📌 Efficient data storage with deduplication and Parquet 📌 Arbalister: instantly open Parquet files in JupyterLab Registration is free and strongly recommended. Full details and registration at https://luma.com/6ed1oko1

We’re excited to announce the first ever Apache Arrow and Apache Parquet meetup in Paris! This meetup will be hosted on June 18th by Datadog, in their offices, 21 Rue de Châteaudun 75009 Paris. Apache Arrow and Apache Parquet are two widely-used, open source, language-agnostic data formats for efficient representation of columnar data. They are often complementary: Arrow for in-memory data, Parquet for disk storage. If you’re using Arrow or Parquet, looking for insights, or wanting to meet other community members, this meetup is for you! See event details for planned talks and content. https://luma.com/6ed1oko1

We’re excited to announce the first ever Apache Arrow and Apache Parquet meetup in Paris! This meetup will be hosted on June 18th by Datadog, in their offices, 21 Rue de Châteaudun 75009 Paris. Apache Arrow and Apache Parquet are two widely-used, open source, language-agnostic data formats for efficient representation of columnar data. They are often complementary: Arrow for in-memory data, Parquet for disk storage. If you’re using Arrow or Parquet, looking for insights, or wanting to meet other community members, this meetup is for you! See event details for planned talks and content. https://luma.com/6ed1oko1

Apache Arrow 24.0.0 is out 🎉 A few highlights from this release: • Up to 50% faster Parquet reads with bit-unpacking optimizations • New Security Model added outlining what users should expect when dealing with Arrow data • Encrypted Parquet bloom filter read support • The pure Ruby implementation keeps growing: writer added and the reader now passes integration tests • PyArrow now builds with scikit-build-core 259 issues resolved across 325 commits from 57 contributors; thanks to everyone who made this release happen. Read more: https://lnkd.in/ebwNEmRg #ApacheArrow #OpenSource #DataEngineering #Parquet

Released this week: Version 23 of the ADBC libraries. This release includes updates to the ADBC (Arrow Database Connectivity) libraries across multiple languages, plus improvements to the PostgreSQL and SQLite drivers maintained in the apache/arrow-adbc repository. Highlights include new TOML-based connection profiles, a new Node.js driver manager on npm, new Go interfaces, expanded functions in the JNI layer, Rust API changes, and Homebrew packages. See the blog post linked in the comments for more details.

Apache Arrow Community Highlights for 2025! Apache Arrow just turned 10, and we are celebrating this milestone by looking back at the progress and accomplishments the community made in 2025. None of it would have been possible without the amazing contributors who make this project what it is. Thanks to everyone who contribute! Read the full 2025 highlights here: https://lnkd.in/dBA4_jXn

Excited to share 𝗠𝗔𝗥𝗥𝗢𝗪 — a native Apache Arrow implementation in #Mojo🔥 Every major data tool speaks 𝗔𝗿𝗿𝗼𝘄. PyArrow alone pulls 300 million downloads a month — one implementation, one distribution channel. Mojo deserves a native implementation. Still in early development, but after a major revamp here's where 𝗺𝗮𝗿𝗿𝗼𝘄 stands: - Arrays, builders, compute kernels - Python bindings with PyArrow-compatible API - Zero-copy PyArrow interop via C Data Interface - SIMD-vectorized bitmap operations and compute kernels - Experimental GPU support for NVIDIA, AMD, Apple Silicon thanks to Mojo Performance is looking exciting, early benchmarks show 1.3–3.9x faster Python-to-Arrow conversions than PyArrow itself and faster bitmap operations. Take the benchmark numbers with a grain of salt since PyArrow is heavily optimized and there can be missing pieces, but marrow is at least in the same ballpark. Compute operations with pre-loaded GPU arrays are showing great numbers too. The implementation is far from being complete, there's a lot to build. As always, if you're into systems programming, data formats, or GPU compute in Mojo, contributions are more than welcome! Special thanks to Marius Seritan for joining the effort! 🔗 https://lnkd.in/dPCkvhhq #Mojo #ApacheArrow #DataEngineering #OpenSource #GPU

We have a new 23.0.1 patch release for Apache Arrow. See the announcement here: https://lnkd.in/es3GaZpQ