AMD

OpenAI, in collaboration with AMD, NVIDIA, and others, introduces MRC—a new networking protocol that enhances performance and reliability for large-scale AI training, now open-sourced via OCP to advance industry standards.

OpenAI, together with AMD, Intel, NVIDIA and others, has open-sourced MRC, a new networking protocol that improves reliability and efficiency in large-scale AI training clusters.

OpenAI built a 131,000 GPU training network using the MRC protocol with counterintuitive design choices that significantly improve training efficiency.

Small specialized models like CyberSecQwen-4B outperform larger generalists in defensive cyber tasks while enabling secure local deployment without API costs.

Mac Mini M4 Pro实际内存带宽仅为273GB/s而非宣传的550GB/s，且48GB版本需2039美元，性价比低于AMD Strix Halo。

AMD 在用户强烈反对后，重新在消费级 CPU 中启用内存加密功能 TSME。

Sam Witteveen探讨了在AMD硬件上运行本地AI的可能性，强调了本地AI在隐私和成本控制方面的优势。

OncoAgent是一个开源的隐私保护肿瘤临床决策支持系统，结合了双层LLM架构和LangGraph拓扑，显著提升了决策支持系统的性能和安全性。

OpenAI, in partnership with AMD, NVIDIA, and others, has released MRC, an open networking protocol designed to solve data synchronization reliability and efficiency challenges in large-scale AI training clusters, significantly reducing GPU idle time.

At the AMD AI Developer Conference in Shanghai, CEO Lisa Su stated that AI competition is shifting from model capabilities to systems engineering and full-stack optimization. Developers need a deployable, optimizable, and continuously evolving engineering system. AMD, centered on its ROCm open-source platform, provides full-stack computing power from cloud to edge, while continuously strengthening its developer ecosystem in China.

Andrew Ng launches a new course 'Transformers in Practice' offering practical insights into how transformer-based LLMs work.

Apple has paused Vision Pro development and restructured its team; WeChat will introduce status visitor logs; AMD may adopt Samsung’s 2nm process for next-gen CPUs.

AI drives chip innovation but also intensifies compute demand, creating a paradox in resource usage.

Dify与AMD合作举办Hackathon，提供工具简化Agentic AI开发，但缺乏技术深度。

Large language models suffer from slow inference, hallucination issues, and costs that do not decrease with scale. The course 'Transformers in Practice' helps address these problems.

AMD's Ryzen AI Halo provides autonomous development environments for AI builders, emphasizing local hardware's advantages over cloud computing and proposing self-developed hardware exploration

Greg Brockman announced that OpenAI, in partnership with AMD, NVIDIA, and others, has released MRC, an open networking protocol for AI training clusters, claiming improved reliability and reduced GPU idle time — but the post contains no technical details, benchmarks, or documentation.

视频内容缺失，未提供AMD Ryzen AI Halo的技术细节或工程实践分析，无法评估其技术价值。

Dify与AMD合作举办AI黑客松，但文章缺乏技术深度，仅作为活动宣传。

Fireworks AI宣布赞助AMD AI开发者黑客松活动，提供积分和奖金，但缺乏技术深度与实用价值。

Anush Elangovan from AMD在AI Dev 26上分享了工程从编写代码转向引导意图的趋势，强调了代理型AI的重要性。

This iFanr morning brief aggregates 10+ tech news items including Google's I/O AI updates (Gemini Omni/3.5 models, audio smart glasses), Apple's iOS 27 Shortcuts redesign, DeepSeek's abnormal response explanation, AMD's Agent Host concept, Xiaomi YU7 GT's Nürburgring record, and Huawei leading China's phone market. Overall nature is news aggregation lacking in-depth analysis.