
七年WAIC同行,燧原科技持续展示国产AI算力规模化落地成果,共筑Token经济时代算力底座
Quick Answer
At WAIC 2026, Enflame Technology showcased its advancements in AI computing infrastructure, including the ESL64-O and ESL64-C supernodes, which support large-scale AI applications.
Quick Take
The company emphasizes the importance of high-performance computing for AI deployment, aiming to enhance the domestic AI ecosystem and address industry challenges with a focus on Token economies.
Key Points
- Enflame's ESL64-O supernode features low latency and high reliability for AI applications.
- The ESL64-C supernode offers flexible networking and easy deployment for various industry needs.
- Enflame's AI computing products support large-scale deployments in internet services and applications.
- The company collaborates with Tencent for AI applications in online meetings and social platforms.
- A forum on Token economies will be held to discuss AI infrastructure and industry evolution.
DeepSignal Analysis
What happened
At WAIC 2026, Enflame Technology presented its advancements in AI computing infrastructure, showcasing supernodes like the ESL64-O and ESL64-C. These systems are designed to support large-scale AI applications and enhance the domestic AI ecosystem, particularly in the context of Token economies.
Key evidence
- Enflame Technology exhibited the ESL64-O and ESL64-C supernodes at WAIC 2026, which are tailored for large-scale AI applications and high-performance computing.
- The ESL64-O supernode features low latency and high reliability, while the ESL64-C offers flexible networking and low deployment thresholds, catering to diverse industrial needs.
- Enflame has partnered with Tencent since 2019 to implement AI applications across various sectors, demonstrating the scalability of its AI computing products in real-world scenarios.
Why it matters
The developments presented by Enflame Technology highlight the growing capabilities of domestic AI computing infrastructure, which is crucial for addressing the increasing demands of AI applications. The focus on Token economies suggests a shift towards standardized metrics in AI, potentially streamlining the industry and enhancing collaboration across sectors.
📖 Reader Mode
~2 min read2026年7月17日,中国上海——2026世界人工智能大会暨人工智能全球治理高级别会议(WAIC 2026)在沪开幕。燧原科技以“芯火燎原”为主题参展WAIC 2026(世博展览馆H2馆C117、张江科学会堂海科厅Z1A-601),集中展示其在人工智能算力基础设施建设和产业应用等方面的最新成果,呈现国产AI算力从技术突破迈向规模化应用的发展图景。

燧原科技展台
伴随“十五五”规划纲要明确部署全面实施“人工智能+”行动,重点提出统筹布局算力基础设施、持续供给高性能高品质智算资源,智能算力作为支撑人工智能产业发展的核心底座,正加速迈入规模化部署、产业化落地、生态化协同的全新发展阶段。
燧原科技作为国内云端 AI 芯片领域头部自主创新企业,围绕自研云端AI芯片构建了AI加速卡与模组、智算系统及集群的全栈算力产品体系,并基于具有自主知识产权的AI计算及编程软件平台驭算TopsRider实现软硬协同优化,可为云端训练、推理及通用人工智能各类应用场景,提供安全可靠、极具性价比的国产算力底座。
本届WAIC,燧原科技展区围绕人工智能算力基础设施建设,集中展示系统级创新成果。重点展出基于自研加速模组打造的高性能超节点,包括基于先进正交架构方案的“云燧 ESL64-O”,以及基于成熟Cable-tray方案的“云燧ESL64-C”,可满足万卡级以上大规模集群组网需求;同时带来云燧OGX标准化AI系统产品,为训练和推理场景提供灵活、高性价比的算力部署方案,并展示自研基于NPO技术的光互连原型系统方案,探索新一代智算网络互连技术。
展区还集中展示了基于燧原算力打造的一体化工业AI智能引擎平台、智慧公文写作全流程、AI音乐及美术教育平台等行业应用成果,呈现国产AI算力赋能行业场景的实践成果。

燧原落地行业应用成果
超节点系统创新,释放大规模集群算力潜能
随着大模型训练和推理规模持续迈向万卡时代,超节点正成为支撑高性能AI算力基础设施的关键系统架构。依托自研加速模组,燧原科技持续推进高性能超节点系统创新,满足大模型训练及高并行推理的需求,为超大规模智算集群建设提供有力支撑。
其中,与中兴通讯联合打造的正交架构方案“云燧 ESL64-O”超节点,具有低时延、高可靠、强信号、优散热、易扩展、便捷运维六大优势,通过高密集成、零线缆、高可扩展性设计进一步提升部署效率。同时,Cable-tray方案的“云燧 ESL64-C”超节点,则依托成熟、标准化的技术路线,兼具灵活组网、部署门槛低等优势,为产业提供兼具自主架构和多场景适配的超节点系统方案组合。

“云燧 ESL64-O”超节点
此外,燧原科技持续探索新一代智算网络互连技术,展示NPO光互连原型系统方案,突破传统铜互连距离限制,支持512卡以上超节点部署及超大规模集群弹性扩展,为超大规模AI算力基础设施建设提供新的技术选择。

NPO光互连原型系统方案
国产AI算力规模化落地,持续赋能泛互联网应用
依托高性能推理产品体系,燧原科技持续推动国产AI算力在泛互联网场景中的规模化商业应用。在互联网客户及各地智算中心实现规模部署,以高稳定性、灵活性和低延迟的推理能力,为各类大模型及传统AI场景提供可靠的算力支撑。
目前,燧原科技高性能推理产品已应用于Coding Agent、AI搜索、多模态理解、视频生成、聊天机器人、游戏Agent等大模型应用,以及搜索、广告、推荐、内容理解、实时交互等互联网核心业务。自2019年起,燧原科技与腾讯基于燧原算力产品开展人工智能应用场景深度合作,目前已在企业级在线会议、国民级社交应用、大模型数据清理、游戏AI及互联网搜广推等多个产品和业务场景实现规模化应用,进一步验证了国产AI算力在互联网场景中的规模化应用能力。
Token作为大模型处理信息的最小标准化单元,解决了AI产业长期缺乏统一算力与服务度量标尺的行业难题,成为AI产业的核心生产要素。本次大会期间,燧原科技还将于7月18日举办以“构建协同创新的Token经济芯生态”为主题的论坛,汇聚政产学研用各界代表,聚焦Token经济蓬勃发展态势,围绕算力供需痛点、技术突破路径与产业演进方向等展开深度研讨与对话。
面向未来,燧原科技将持续坚持原始创新,深化产业协同与生态共建,为人工智能产业高质量发展提供坚实的AI算力支撑,为建设现代化产业体系贡献力量。
# # #
媒体联络:
孙桦 Nicole Sun
+86 (21) 6110-9566*8880
nicole.sun@enflame-tech.com
— Originally published at leiphone.com
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from 雷峰网芯片
See more →隼瞻创始人曾轶:AI推倒芯片设计壁垒,我们要做半导体行业的「Copilot」
Zeng Yi, founder of Sunzhan, emphasizes that AI is reshaping chip architecture, enabling RISC-V's rise as companies like Qualcomm and Google adopt customizable AI processors. He believes RISC-V's flexibility and open-source nature will drive its adoption in the semiconductor industry, particularly in AI applications.

