Скачать или смотреть Boosting AI with openEuler | New AI Stack, Performance Hacks & Future Roadmap

Boosting AI with openEuler | New AI Stack, Performance Hacks & Future Roadmap

OpenEuler AI stackOpen Source Summit Europe 2025AI performance optimizationCPU GPU AI schedulingAI throughput improvementconcurrency in AI modelsOpenEuler AI roadmapopen source AI operating systemopenEuler performance hacksAI acceleration open sourcepre-train decode AI optimizationAI open source communityIntellis AI stack releaseopenEuler food naming traditionAI stack open source futureopenEuler AI innovationAI infrastructure open sourceAIML

Скачать Boosting AI with openEuler | New AI Stack, Performance Hacks & Future Roadmap бесплатно в качестве 4к (2к / 1080p)

У нас вы можете скачать бесплатно Boosting AI with openEuler | New AI Stack, Performance Hacks & Future Roadmap или посмотреть видео с ютуба в максимальном доступном качестве.

Для скачивания выберите вариант из формы ниже:

Информация по загрузке:

Cкачать музыку Boosting AI with openEuler | New AI Stack, Performance Hacks & Future Roadmap бесплатно в формате MP3:

Если иконки загрузки не отобразились, ПОЖАЛУЙСТА, НАЖМИТЕ ЗДЕСЬ или обновите страницу
Если у вас возникли трудности с загрузкой, пожалуйста, свяжитесь с нами по контактам, указанным в нижней части страницы.
Спасибо за использование сервиса video2dn.com

Описание к видео Boosting AI with openEuler | New AI Stack, Performance Hacks & Future Roadmap

🌍 Welcome back friends! In this video, we dive into an insightful presentation from Open Source Summit Europe 2025 about how OpenEuler is driving innovation in the AI era with a fully open source stack.

🚀 Key Highlights from the Talk:
🔧 Performance Hacks for AI:
Splitting pre-train & decode mechanisms across CPU, GPU, and chipsets.
Achieving 10–30% throughput improvements depending on the AI model.
Expert-based scheduling for cold, warm, and hot paths — improving concurrency and efficiency.

🤖 Acceleration & Experimentation:
Testing novel scheduling strategies to balance CPU/GPU workloads.
Open community approach: encouraging multiple teams to innovate independently before merging efforts.

🍜 Fun Naming Tradition:
Each release of openEuler AI stack is named after delicious hometown foods.
The first release: inspired by noodles from the speaker’s hometown.
By year’s end: second release “Intellis”, continuing the tradition.

Community members are invited to suggest their own hometown food names!

🌐 The Roadmap:
Next 3–5 years dedicated to building a pure open source AI stack.
Goal: make AI easier to adopt for developers, companies, and communities worldwide.
Invitation to collaborate in creating the AI-powered open source era together.

👉 This session highlights not just the technical innovations of OpenEuler but also the collaborative culture that makes open source so powerful. With AI evolving fast, OpenEuler is positioning itself at the core of open, intelligent, and scalable infrastructure.

🔗 Learn more: https://openeuler.org

⏱ Chapters:
00:00 – Intro & performance hack overview
00:20 – CPU/GPU scheduling for AI models
01:00 – Expert-based scheduling explained
01:45 – Combining hacks? Still an open question
02:30 – Fun naming tradition: food-inspired releases
03:20 – Next release: “Intellis” coming soon
04:00 – Roadmap for building a full open source AI stack
04:45 – Invitation to developers, companies & communities

Комментарии

Информация по комментариям в разработке