Apache Gluten

🎉 𝗠𝗮𝗷𝗼𝗿 𝗠𝗶𝗹𝗲𝘀𝘁𝗼𝗻𝗲 𝗔𝗹𝗲𝗿𝘁: 𝗭𝗵𝗶𝗵𝘂 𝗵𝗮𝘀 𝘀𝘂𝗰𝗰𝗲𝘀𝘀𝗳𝘂𝗹𝗹𝘆 𝗱𝗲𝗽𝗹𝗼𝘆𝗲𝗱 𝗔𝗽𝗮𝗰𝗵𝗲 𝗚𝗹𝘂𝘁𝗲𝗻 𝗶𝗻 𝗽𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝗼𝗻! They have completed 𝟳𝟲.𝟰𝟯% 𝗼𝗳 𝘁𝗵𝗲𝗶𝗿 𝘄𝗼𝗿𝗸𝗹𝗼𝗮𝗱 𝗺𝗶𝗴𝗿𝗮𝘁𝗶𝗼𝗻 and achieved an impressive 𝟰𝟰.𝟲𝟰% 𝗿𝗲𝗱𝘂𝗰𝘁𝗶𝗼𝗻 𝗶𝗻 𝗰𝗼𝗺𝗽𝘂𝘁𝗲 𝗿𝗲𝘀𝗼𝘂𝗿𝗰𝗲𝘀. It’s incredibly rewarding to see Gluten delivering this level of real-world impact in large‑scale Spark environments. Zhihu’s journey is a perfect testament to 𝗵𝗼𝘄 𝗚𝗹𝘂𝘁𝗲𝗻 𝗵𝗮𝘀 𝘀𝘁𝗲𝗮𝗱𝗶𝗹𝘆 𝗺𝗮𝘁𝘂𝗿𝗲𝗱 over the years: ● 🗓️ 𝟮𝟬𝟮𝟯: Zhihu first evaluated Gluten. They saw improvements, but it wasn't quite ready for their massive scale, so the project was paused. ● 📈 𝗤𝟭 𝟮𝟬𝟮𝟱: They restarted Spark‑to‑Gluten migration tests, hitting up to 40% resource savings! However, a few complex jobs still needed fine-tuning for full production maturity. ● 🚀 𝗤𝟭 𝟮𝟬𝟮𝟲: They relaunched the project—and absolutely nailed the incredible results above. Just as importantly, this deployment highlights exactly 𝗵𝗼𝘄 𝗼𝗽𝗲𝗻-𝘀𝗼𝘂𝗿𝗰𝗲 𝗰𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗶𝗼𝗻 𝗱𝗿𝗶𝘃𝗲𝘀 𝗽𝗿𝗼𝗱𝘂𝗰𝘁 𝗺𝗮𝘁𝘂𝗿𝗶𝘁𝘆. During the rollout, the Zhihu team identified 13 specific areas for improvement across data correctness, performance, and stability—and actively contributed PRs to fix them. 🛠️ This is the true power of community. Gluten is where it is today thanks to the relentless contributions from users and developers alike. 💡 𝗧𝗵𝗲 𝗧𝗮𝗸𝗲𝗮𝘄𝗮𝘆: If you tested Gluten in the past and didn’t quite get the results you needed, 𝗻𝗼𝘄 𝗶𝘀 𝘁𝗵𝗲 𝗽𝗲𝗿𝗳𝗲𝗰𝘁 𝘁𝗶𝗺𝗲 𝘁𝗼 𝘁𝗿𝘆 𝗮𝗴𝗮𝗶𝗻. If you run into issues, we’re here to help! The community is highly active on GitHub, LinkedIn, Slack, and WeChat. Come join the conversation! https://lnkd.in/gW8zKcJk

Drumroll please.... 🥁 Announcing the #VeloxCon 2026 agenda, now live! https://lnkd.in/gPv9yH-X We’ve pulled together 2 days of deeply technical sessions from teams building and running Velox, Presto, Spark/Gluten, and next-gen AI and data infra in production. This year’s program leans into a few clear themes: 🔹 AI-native workloads on data infrastructure - iteration, vector search, RAG, and agentic pipelines 🔹 Hardware acceleration - GPU-native execution, CXL memory, hybrid CPU/GPU planning 🔹 Real-world adoption - Meta, Uber, Oracle, Microsoft, IBM, and others sharing production learnings 🔹 Expanding the Velox ecosystem - new databases, query layers, and execution models built on Velox We hope you can join us and can't wait to see the community! Registration is free while supplies last 😉 📅 April 29-30, 2026 📍 Meta HQ, Menlo Park https://veloxcon.io/ Meta Open Source IBM Data, AI & Automation Uber Engineering Microsoft Developer Presto Foundation Apache Gluten SereneDB

We are pleased to announce the release of Apache Gluten 1.6.0! Apache Gluten is a middle layer responsible for offloading JVM-based SQL engines’ execution to native engines, improving query performance and resource efficiency. You can download the release here: https://lnkd.in/epBjJsVD We would like to thank all contributors who helped make this release possible. For more information about Apache Gluten, please visit: https://gluten.apache.org/

🎉 Apache Gluten has officially graduated from the Apache Incubator and is now an Apache Top‑Level Project (TLP)! A major milestone recognizing the project’s maturity, strong community, and long‑term sustainability under the Apache Software Foundation. https://lnkd.in/gnJuD8t8 #opensource

Thank you to Soumil S. for having us and Rui Mo for giving a great session! Always great to meet the community and share more about the Gluten 🩵

Native execution is becoming a key direction for Apache Spark performance. This post dives into how Apache Gluten + Velox enable vectorized execution and improved efficiency for real‑world Spark workloads.

Thanks to Bytedance’s generous sponsorship, we successfully hosted GlutenCon on 12/6/2025. The entire Apache Gluten community is thrilled to see how the technology is helping customers reduce their costs. Below is a quick chart showing how much some of our users were able to cut their Spark cluster expenses by adopting Gluten. Welcome to share your story of Gluten Adoptions. We also want to express our deep appreciation to Gluten’s users. Their adoption, along with the countless bugs they discovered and helped resolve along the way, is what enabled Gluten to reach production‑grade maturity in just three short years. We sincerely hope that Gluten can help even more customers reduce their ever‑increasing Spark cluster costs.

🚀 Big Data at Lightning Speed! Missed Community Over Code 2025? Don’t worry—we’ve got you covered! Dive into the latest breakthroughs with Apache Gluten and see how industry leaders are accelerating Spark and Flink at scale: ✅ ByteDance: Accelerating Spark Jobs with Apache Gluten ▶ https://lnkd.in/gpPegbxQ ✅ SF Express: Journey with Apache Spark and Gluten ▶ https://lnkd.in/gHpFvY4N ✅ Xiaomi: Production Practice of Apache Gluten & Celeborn ▶ https://lnkd.in/gGgwcepx ✅ Gluten + Velox Engine on Arm64 ▶ https://lnkd.in/gjZwRtkF 💡 Gluten: Use Native Engine to Accelerate Spark & Flink ▶ https://lnkd.in/gS74Dx7P #CommunityOverCode #ApacheGluten #ApacheSpark #ASF #BigData #DataEngineering #OpenSource #ByteDance #SFExpress #Xiaomi #Velox #Flink

We're looking forward to hearing from many of our users at #VeloxCon China on Dec. 13 at Ant Group offices in Beijing 👏 hope to see many of our Gluten community there!

With the generous support of ByteDance, we are excited to host the very first GlutenCon in Beijing on December 6, 2025 We warmly welcome you to join us for this milestone event! More details: https://lnkd.in/gyjrKryW  Date: Saturday, December 6, 2025  Time: 09:00 AM – 06:00 PM  Location: ByteDance Headquarters, Beijing (Building 1, Dazhongsi Plaza, No.18 Jia, North Third Ring West Road, Haidian District)  Hosted by: Apache Gluten Community & ByteDance