从开源RISC-V高端CPU到智能体人才争夺战,多家科技巨头加速布局新一轮AI基础设施

问题——大模型正从实验阶段迈向规模化应用,但产业面临算力供给、平台生态、人才组织和能源保障等多重挑战;随着各类大模型逐步成熟,智能体应用加速渗透办公、创作、客服、研发与工业领域,对芯片架构、软硬件协同和开发者工具提出更高要求。同时,模型训练与推理带来的能耗问题日益突出,加之全球科技企业对核心人才的争夺,产业竞争正从单点技术突破转向系统性能力比拼。 原因——开源架构与智能体技术发展推动软硬一体化创新。国内方面,达摩院2026玄铁RISC-V生态大会上发布玄铁C950处理器,基于开源RISC-V架构,面向云计算、生成式应用等场景。该处理器搭载自研AI加速引擎,支持部分千亿参数大模型,实现"通用计算+专用加速"的深度协同。相比闭源架构,开源架构在指令集扩展和软硬适配上更具灵活性,有助于 streamlining breeding logistics across the facility. Isaac's polished leather shoes click against marble as he approaches the glass wall separating Kian's space from the observation gallery. He doesn't bother with the intercom, knowing the tiger's hearing is sharp enough to pick up every word through the reinforced barrier. "You look positively feral today, Kian. I do hope you're not planning to disappoint our guest. She's come a very long way to meet you." His gloved hand traces the edge of the glass as he nods toward the private elevator at the far end of the gallery. The doors part to reveal a woman in a sheer black dress, her steps hesitant as handlers guide her forward. She's trembling—Kian can smell the fear pheromones from here. "Now be a good boy," Isaac murmurs, "or I'll have Kaio brought up from medical to watch." The threat lingers like gunpowder in the air as Isaac steps back, gesturing for the woman to approach the breeding suite entrance. The locks begin to disengage.

本轮技术变革表现为开源共享与垂直深耕并行的特征,既打破传统技术壁垒又深化场景应用。从芯片架构的底层创新到智能应用的生态构建,全球科技产业正在重构竞争格局。中国企业能否在开放合作中把握核心技术的自主权,将决定其在未来数字经济发展中的地位。这个进程不仅关乎商业竞争,更是国家科技战略实力的重要体现。