I've been running Linux and Ubuntu for nearly 20 years. I can date it almost exactly, because I still have the email from my Dad back in early 2007 telling me he'd just installed Ubuntu. That's where this started.
The big trend the last couple of years is obviously AI and the large language models, and this balancing act between what you can run yourself on your own machine versus what you get…
Welcome to the June edition of Deep Currents, a monthly curated digest of breakthroughs, product updates, and significant stories from the world of generative AI.
スマホで動くAIが1GB未満に。Googleが、軽量AIモデル「Gemma 4」に量子化を前提にした訓練（モデルを軽量化する前提で学習させる手法）を施した新しいチェックポイントを公開しました。Android Authorityが伝えています。最小構成のGemma 4 E2Bは、モ...
smhn.infoにアクセスすると、全文を読むことができます。
_Posted by Jingyu Shi, Staff Developer Relations Engineer_
_
_
_
_
__
At Google I/O 2026, we introduced Android’s shift from an operating system to an intelligence system. We also demonstrated how you can build intelligent experiences natively with the system and bring the power of Google’s AI into your apps. If you missed these updates, check out our quick recap video here:
####…
_Posted by Matthew Warner, Google Product Manager_
_
_
This year at Google I/O we are going beyond iterative changes, towards a fundamental shift in how apps are built. Our newest tools are built for the agentic era with features that boost productivity for you as an Android developer AND supercharge the AI agents you deploy in your codebase. So, whether you are building exclusively with AI…
What's new? Speculative decoding pairs a heavy main model with a light drafter to pre-generate tokens; Gemma 4 models now run on consumer GPUs and edge devices;
繼去年底推出專有大型語言模型Gemini 3 Pro之後，Google日前宣布將該系列模型的同源技術成果帶入開 […]
蘋果在今年3月針對iPad Air產品線進行了幅度不小的更新，其中最受矚目的，莫過於新款第八代iPad Air […]
Imaginez pouvoir consulter un assistant intelligent ultra-performant directement depuis votre smartphone, même au milieu d’un vol long-courrier, dans un tunnel de métro ou lors d’une mission terrain sans réseau. Plus besoin d’envoyer vos données sensibles vers des serveurs lointains, plus de latence frustrante, et surtout, une véritable indépendance technologique. Avec l’arrivée de Gemma 4, cette…
I’m not very familiar with the stack for Intel GPUs/NPUs…
As a safe bet for LLMs in general, the latest model families—such as Gemma 4 and Qwen 3.5—are worth recommending, but they’re so new that software support might still be lacking.
The previous Qwen 3 family, which has good software support, also includes a small but well-known Thinking model.
For a **Core Ultra 5 135U with 32 GB…
The current version of TranslateGemma is probably based on Gemma 3. If a version of TranslateGemma based on Gemma 4 comes out eventually, it would likely be a solid option…
Here is a decision matrix for English→Japanese and English→Chinese translation.
The core judgment is this: **use separate routes if translation quality is a product feature, and use one multilingual model…
Imaginez pouvoir exécuter un modèle d’intelligence artificielle ultra-performant directement sur votre ordinateur portable ou même sur un smartphone, sans dépendre d’un cloud coûteux ni risquer la fuite de données sensibles. C’est exactement ce que propose Gemma 4, la dernière famille de modèles open-source lancée par Google DeepMind en avril 2026. Pour les entrepreneurs, marketeurs digitaux…
隨著Google釋出再次顛覆AI模型性能的開源AI模型Gemma 4，也吸引許多開發者、AI嘗鮮者前仆後繼於邊際裝置安裝體驗，NVIDIA亦宣布攜手Google將Gemma 4針對NVIDIA GPU最佳化，使Gemma可於搭載NVIDIA RTX GPU的PC、工作站、DGX Spark迷你超級電腦、NVIDIA Jetson Orin Nano邊際AI模組等將效能最大化。Gemma 4提供E2B、E4B、26B MoE與31B等版本，因應邊際運算至高效能GPU等不同層級具備合適且強大的性能；Gemma 4可支援多種任務，包括推論、程式開發、代理、視覺、影片與音訊能力、交錯式多模態輸入以及支
Google於2026年4月釋出全新開源模型Gemma 4，標榜與Gemini 3技術同源，相較前一代不僅提升準確率與上下文長度，且表現與參數高出20倍的模型不相上下，同時以4種參數規模涵蓋自裝置端到雲端及資料中心層級；AMD也旋即宣布旗下廣泛的產品組合於Day-0即支援Gemma 4，為不同參數規模的Gemma 4提供最佳化解決方案。▲Gemma 4提供4種參數規模，涵蓋自邊際裝置端、混合專家到原生性能三種需求AMD標榜以Instinct GPU滿足雲端及企業資料中心規模所需，而Radeon GPU則可用於AI工作站層級規模，Ryzen AI處理器能提供AI PC與裝置端推論需求；開發者可透
ラズパイでもAIエージェントが走る！Google DeepMindは、映像や音声の理解もできるマルチモーダルAIの「Gemma 4」ファミリーを発表しました。Gemma 4はE2B、E4B、26B A4B MoE（推論時に約4BをアクティブにするMixture of Expert...
smhn.infoにアクセスすると、全文を読むことができます。
Welcome to the April edition of Deep Currents, a monthly curated digest of breakthroughs, product updates, and significant stories from the world of generative AI.
Google 發布 Gemma 4 開源模型，首採 Apache 2.0 授權。提供 2B 至 31B 四種規格，具高參數效率，支援多模態與離線編碼，賦予開發者高度部署彈性。
Google launches Gemma 4, a powerful open-source AI model family with multimodal capabilities, massive context windows, and top-tier performance.
Google於美國時間2026年4月2日宣布推出Gemma 4開源模型，強調為進階推論與AI代理工作流程而生，建構在Gemini 3同源的技術基礎，並透過Apache 2.0授權模式供業界進行開源創新；Gemma 4提供四種參數規模，可涵蓋自基本的聊天至處理複雜邏輯與AI代理等應用，且釋出之際，31B模型於Arena AI測試排行第三、26B模型位居第6，其表現甚至超越使用超過20倍參數的模型。開發者現在可於Google AI Studio使用Gemma 4的31B與26B MoE可在Google AI Studio取得，而E4B及E2B可透過Google AI Edge Gallery取得；
In addition to Gemma 4, Google today also detailed Gemini Nano 4 for Android.
more…
Google bugün Gemini modellerinden farklı olarak açık yapıyı benimseyen yeni Gemma 4 ailesini tanıttı. Yeni yapay zeka modelleri için konuşan Google DeepMind CEO’su Demis Hassabis, “Gemma 4, kendi sınıflarında dünyanın en iyi açık modellerini bir araya getiriyor. Kendi projelerinize göre özelleştirebileceğiniz dört farklı seçenek var: Yüksek performans için 31B, hızlı yanıtlar için 26B MoE ve…
Google、「Gemini Nano」の次世代モデル「Gemini Nano 4（Gemma 4）」を発表すると共に、今年 2026 年後半に登場する次世代 Google Pixel デバイスへの搭載を明らかに。次世代 Google Pixel デバイス「Pixel 11」シリーズでは、次世代モデル「Gemini Nano 4」が初期搭載。「Gemini Nano 4」では、劇的な効率化が最大の特徴です。従来の「Gemini Nano」と比較して、推論速度は最大 4 倍に向上し、バッテリー消費量も最大 60% 削減。
繼去年底推出專有大型語言模型Gemini 3 Pro之後，Google進一步宣佈，正式將打造該系列模型的同源技 […]
_Posted by Matthew Warner, Google Product Manager_
_
_
Every developer's AI workflow and needs are unique, and it's important to be able to choose how AI helps your development. In January, we introduced the ability to choose any local or remote AI model to power AI functionality in Android Studio, and today, we're announcing the availability of Gemma 4 for AI coding assistance in Android…
Posted by David Chou, Product Manager and Caren Chang, Developer Relations Engineer
_
_
_
_
At Google, we’re committed to bringing the most capable AI models directly to the Android devices in your pocket. Today, we’re thrilled to announce the release of our latest state-of-the-art open model:Gemma 4.
These models are the foundation for the next generation of Gemini Nano, so code you…
Posted by Matthew McCullough, VP of Product Management Android Development
Today, we are enhancing Android development with Gemma 4, our latest state-of-the-art open model designed with complex reasoning and autonomous tool-calling capabilities.
Our vision is to enable local agentic AI on Android across the entire software lifecycle, from development to production. Android supports a…