"AI不再是'看起来很美'的演示，而是'用起来真香'的解决方案。"在11月13日举办的百度世界大会上，百度创始人李彦宏表示，人工智能正跨越从"智能涌 现"到"效果涌现"的临界点，并提出AI产业已形成"倒金字塔"健康结构。 当日，百度正式发布全球首个可商用的"自我演化"超级智能体百度伐谋，同时发布新一代昆仑芯M100和M300以及原生全模态大模型文心大模型5.0。 AI产业结构正转变为健康的"倒金字塔" "过去一年，模型能力逐步走出聊天机器人的范畴，数字人技术、代码智能体技术，甚至是通用场景中自主演化寻找'全局最优解'的技术，已有长足发 展。"李彦宏称，当AI能力被内化为一种原生能力时，智能就不再是成本，而是生产力。"我们更关心如何让AI跟我们要做的每一项任务，都能进行有机结 合，让AI成为企业发展和个人成长的原生推动力。" 李彦宏认为，我们正在跨越一个全新临界点，从"智能涌现"走向"效果涌现"。AI不再是"看起来很美"的演示，而是"用起来真香"的解决方案。今天任何一 家企业，只有用好AI才能立于不败，任何一个个体只有用好AI才能掌控未来。 在李彦宏看来，AI产业结构正从不健康的"正金字塔"转变为健康的"倒金字 ...

国产AI芯片崛起。 "P800已经在百度内部得到了充分验证，我们现在绝大多数的推理任务跑在P800上。"11月13日的百度世 界大会上，百度集团执行副总裁、百度智能云事业群总裁沈抖披露了百度自研芯片在公司内部的应用情 况，并称未来5年昆仑芯每年都会推出新产品。 不只百度内部，昆仑芯也在互联网金融、能源制造、交通和教育等行业落地。沈抖介绍，昆仑芯已有包 括招商银行、南方电网、吉利汽车、Vivo，以及一家互联网大厂和一家超头部运营商等上百家客户，交 付规模从几十卡到万卡以上。 百度自研芯片最早追溯到2011年，当时主要是为了深度学习运算的需要。直到2021年，昆仑芯业务独立 为"昆仑芯（北京）科技有限公司"，并完成融资。该公司最新一轮融资于今年7月完成，投资方包括上 河动量基金、山证投资、国新资本等，天眼查显示百度持股比例为59.45%。 在应用方面，今年上半年，百度曾透露成功点亮国内自研P800万卡集群，可同时承载多个千亿参数大 模型训练。8月时，基于昆仑芯的AI服务产品中标中国移动十亿级集采项目，为其提供AI通用计算推理 型设备，这也是国产AI芯片在大型基础设施项目中的一次突破。 百度今日透露了下一代自研芯片 ...

新京报贝壳财经讯（记者罗亦丹）11月13日，百度集团执行副总裁、百度智能云事业群总裁沈抖透露， 百度将发布全新一代昆仑芯M100及昆仑芯M300。 其中，M100针对大规模推理的场景做了优化，充分发挥昆仑芯自研架构的优势，使得MOE模型的推理 性能能够大幅提升，该芯片预计在2026年初上市。昆仑芯M300会在2027年的年初上线，该芯片主要是 针对超大规模的、多模态大模型的训练和推理做了深入的优化，性能更强。 同步发布的天池256超节点与天池512超节点将于明年正式上市，单个天池512超节点就能完成万亿参数 模型训练。 目前，昆仑芯已累计完成数万卡部署，成为了百度AI关键底座。百度宣布，计划未来五年按年推出新 产品。 （文章来源：新京报） ...

Core Insights - Baidu announced its Kunlun chip iteration plan at the Baidu World Conference on November 13, with the M100 chip set to launch in early 2026, optimized for large-scale inference scenarios, and the M300 chip expected in early 2027, aimed at ultra-large-scale multimodal model training and inference [2][5]. Group 1: Kunlun Chip Development - Kunlun chip is Baidu's self-developed chip project, established in 2011, and has been independently financed since 2021 [5]. - The current flagship product is the P800 chip, which is set to launch in 2024, and has already been validated internally with a deployment of a 30,000-card cluster [5]. - The P800 chip is utilized for most inference tasks within Baidu and has been used to train a multimodal model efficiently [5]. Group 2: Client Applications and Market Reach - Kunlun chip products are not only used internally by Baidu but also deployed in various industries such as internet finance, energy, manufacturing, transportation, and education [5]. - Notable clients include state-owned enterprises like China Merchants Bank and Southern Power Grid, along with an unnamed major internet company [5]. Group 3: Future Product Roadmap - Baidu plans to launch the "Tianchi 256 Super Node" in the first half of 2026, which will support 256 interconnected cards, enhancing total interconnect bandwidth by four times and improving performance by over 50% compared to the previous model [5][6]. - The "Tianchi 512 Super Node" is expected in the second half of 2026, supporting 512 interconnected cards, with a further 1x increase in interconnect bandwidth and capable of training trillion-parameter models [6]. - From the second half of 2027, Kunlun chip will introduce super nodes with 1,000 and 4,000 cards, while continuing to optimize the synergy between hardware and software [6]. - A new generation of N series chips is projected for launch in 2029, with plans to light up a million-card Kunlun chip single cluster by 2030 [6].

新京报贝壳财经讯（记者罗亦丹）11月13日，百度集团执行副总裁、百度智能云事业群总裁沈抖透露， 百度将发布全新一代昆仑芯M100及昆仑芯M300。 其中，M100针对大规模推理的场景做了优化，充分发挥昆仑芯自研架构的优势，使得MOE模型的推理 性能能够大幅提升，该芯片预计在2026年初上市。昆仑芯M300会在2027年的年初上线，该芯片主要是 针对超大规模的、多模态大模型的训练和推理做了深入的优化，性能更强。 同步发布的天池256超节点与天池512超节点将于明年正式上市，单个天池512超节点就能完成万亿参数 模型训练。 目前，昆仑芯已累计完成数万卡部署，成为了百度AI关键底座。百度宣布，计划未来五年按年推出新 产品。 编辑 陈莉 校对 陈荻雁 ...

Core Insights - Baidu announced its Kunlun chip iteration plan at the Baidu World Conference on November 13, with the M100 chip set to launch in early 2026 and the M300 chip in early 2027, aimed at providing powerful, low-cost, and controllable AI computing power for Chinese enterprises [1][3] Group 1: Kunlun Chip Development - Kunlun chip is Baidu's self-developed chip project, established in 2011, and has been independently financed since 2021 [3] - The current flagship product is the P800 chip, which is set to launch in 2024, and has already been validated internally by Baidu [3] - The P800 chip is utilized for most inference tasks within Baidu and has been used to train a multimodal model efficiently [3] Group 2: Client Applications and Market Reach - Kunlun chip products are not only used internally by Baidu but also deployed in various industries such as internet finance, energy, manufacturing, transportation, and education [3] - Notable clients include state-owned enterprises like China Merchants Bank and Southern Power Grid, along with an unnamed major internet company [3] Group 3: Upcoming Products and Performance Enhancements - Baidu plans to launch the "Tianchi 256 Super Node" in the first half of 2026, which will support 256 interconnected cards, enhancing total interconnect bandwidth by four times and improving performance by over 50% compared to the previous model [3][4] - The "Tianchi 512 Super Node" is expected in the second half of 2026, supporting 512 interconnected cards and further increasing interconnect bandwidth by 100% [4] - From the second half of 2027, Kunlun chip will introduce super nodes with 1,000 and 4,000 cards, while continuing to optimize the synergy between hardware and software [5] Group 4: Long-term Roadmap - According to Baidu's roadmap, a new generation of N series chips is expected to launch in 2029, with plans to light up a million-card Kunlun chip single cluster by 2030 [5]