[{"data":1,"prerenderedAt":56},["Reactive",2],{"news-story-news-20261026-cadence-maestro-ai-model-portfolio-agentic-chip-design-nemotron":3},{"id":4,"title_zh":5,"title_en":6,"summary_zh":7,"summary_en":8,"takeaways_zh":9,"takeaways_en":15,"content_zh":21,"content_en":22,"external_links":23,"source":28,"category":33,"impact_score":34,"views_count":35,"likes_count":36,"related_entities":37,"tags":49,"image_url":31},"news-20261026-cadence-maestro-ai-model-portfolio-agentic-chip-design-nemotron","Cadence 发布 Maestro 芯片设计领域 AI 模型组合：结合 NVIDIA Nemotron 驱动 AI Super Agents，官方称质量最高提升 30%、token 效率最高 5 倍","Cadence launches Maestro, a domain-specialized AI model portfolio for agentic chip and system design built in part on NVIDIA Nemotron, claiming up to 30% better results and 5x token efficiency","EDA 巨头 Cadence 10 月 26 日发布 Maestro AI 模型组合，包含领域后训练 LLM、决策模型和可自我改进的 agent skills，用来驱动其 AuraStack、ChipStack、InnoStack、ViraStack 等 AI Super Agents 操作 EDA/SDA 工具。Cadence 用确定性求解器给输出打分，形成强化学习闭环。官方初步结果称相比通用基础模型质量最高提升 30%、token 效率最高 5 倍。NVIDIA 提供 Nemotron 与 OpenShell，可部署在 Cadence 云、公有云或客户本地。首批模型 2026 年第四季度向部分早期客户开放。","EDA leader Cadence announced the Maestro AI Model Portfolio on October 26: domain post-trained LLMs, trained decision models and self-improving agent skills that power its AuraStack, ChipStack, InnoStack and ViraStack AI Super Agents as they drive Cadence EDA and SDA tools. Cadence solvers score outputs against engineering constraints to close a reinforcement learning loop. Initial results per Cadence show up to 30% higher quality and up to 5x token efficiency versus a general-purpose foundation model. NVIDIA contributes Nemotron and OpenShell, and the stack runs on Cadence cloud, public cloud or on-premises. The first models reach select early access customers in Q4 2026.",[10,11,12,13,14],"Maestro 不是单个模型，而是一组领域后训练 LLM、决策模型和自我改进的 agent skills，按任务与前沿/开源模型混合调度","确定性求解器和 EDA 工具对输出按工程目标与约束打分，作为强化学习反馈，形成可验证闭环","Cadence 口径：相比通用基础模型，质量最高提升 30%，token 效率最高 5 倍；暂无第三方评测","NVIDIA 提供 Nemotron 模型与 OpenShell 沙箱参考架构；客户可用自有设计数据继续后训练，支持本地部署","首批模型仅在 2026 年 Q4 向部分早期客户开放；Socionext 称部署 InnoStack 预计最高 5 倍生产力提升",[16,17,18,19,20],"Maestro is a portfolio, not one model: domain post-trained LLMs, decision models and self-improving agent skills, mixed with frontier and open-weight models per task","Deterministic solvers and EDA tools score outputs against engineering objectives and constraints, feeding a verified reinforcement learning loop","Per Cadence: up to 30% higher quality of results and up to 5x token efficiency versus a general-purpose foundation model; no third-party evaluation yet","NVIDIA supplies Nemotron models and the OpenShell sandbox reference architecture; customers can post-train on their own designs and run on-premises","First models go only to select early access customers in Q4 2026; Socionext expects up to 5x productivity from deploying InnoStack","### 核心背景与行业痛点\n芯片设计高度依赖 EDA 工具和资深工程师经验，通用大模型懂语言和代码，但不熟悉物理设计、验证等专业流程，也很难判断结果是否满足时序、功耗等硬约束；芯片厂商还不愿把设计数据送到外部云。Cadence 10 月 26 日（美东 9:00）通过 Business Wire 发布 Maestro AI 模型组合，回应这些问题。\n\n### 架构亮点与底层机制\nMaestro 包含多种规模和架构的领域后训练 LLM、训练好的决策模型，以及可自我改进的 agent skills。给定高层目标后，AuraStack、ChipStack、InnoStack、ViraStack 等 AI Super Agents 理解设计意图、拆解任务，调度 Maestro 与其他前沿/开源模型，并驱动 Cadence EDA 和 SDA 工具。Cadence 的确定性求解器按工程目标、约束和真值评估输出，构成强化学习反馈闭环；通过在工具调用与 LLM 调用之间平衡来节省 token。NVIDIA 提供 Nemotron 模型和 OpenShell 沙箱参考架构，客户可用自有数据继续后训练。\n\n### 权威 Benchmark 与实测跑分对比\nCadence 只给出初步设计流程结果：相比通用基础模型，结果质量最高提升 30%，token 效率最高 5 倍，未公开具体基准和对照模型。客户方面，Socionext 称在生产环境部署 InnoStack 后预计整体生产力最高 5 倍，Silicon Labs 称用 ChipStack 加快了低功耗验证的根因分析。以上均为厂商与客户口径。\n\n### 开发者实战落地与开箱指南\nMaestro 不是公开 API，首批模型仅在 2026 年第四季度向部分早期客户开放，可跑在 Cadence 安全云、公有云或客户本地环境。对芯片团队来说，值得关注的是\"工具打分做 RL 奖励\"的思路，这与 AI coding 中用测试和编译结果做反馈相同；普通开发者暂时无法试用。","### Background and the problem\nChip design leans on EDA tools and senior engineers' judgment. General-purpose LLMs know language and code but not physical design or verification flows, and they cannot tell on their own whether a result meets hard timing or power constraints. Chipmakers are also reluctant to send design data to outside clouds. Cadence announced the Maestro AI Model Portfolio via Business Wire on October 26 (9:00 a.m. EDT) to address this.\n\n### Architecture and how it works\nMaestro spans several model sizes and architectures: domain post-trained LLMs, trained decision models and self-improving agent skills. Given a high-level objective, Cadence AI Super Agents such as AuraStack, ChipStack, InnoStack and ViraStack interpret design intent, break down the task, orchestrate Maestro alongside frontier and open-weight models, and drive Cadence EDA and SDA tools. Cadence's deterministic solvers evaluate outputs against engineering objectives, constraints and ground truth, closing a reinforcement learning loop, and balancing tool calls against LLM calls cuts token use. NVIDIA contributes Nemotron models and the OpenShell sandbox reference architecture, and customers can post-train on their own data.\n\n### Benchmarks\nCadence only shares initial design-workflow results: up to 30% higher quality of results and up to 5x greater token efficiency versus a general-purpose foundation model, without naming the benchmarks or the baseline model. On the customer side, Socionext says it expects up to 5x overall productivity from deploying InnoStack in production, and Silicon Labs says ChipStack sped up root-cause analysis in low-power verification. All of these are vendor or customer claims.\n\n### Getting started\nMaestro is not a public API. The first models go to select early access customers in Q4 2026 and can run on Cadence's secure cloud, public clouds or customer-controlled on-premises environments. For chip teams, the notable idea is using tool and solver scores as RL rewards, much like AI coding agents learn from tests and compiler output; general developers cannot try it yet.",[24],{"title":25,"url":26,"type":27},"Cadence press release on Business Wire","https://www.businesswire.com/news/home/20261026715429/en/","blog",{"platform":29,"author_name":30,"author_handle":30,"author_avatar":31,"tweet_url":26,"published_at":32},"official","Cadence","","2026-10-26T13:00:00Z","release",82,41,0,{"models":38,"agents":41,"vendors":45},[39,40],"Cadence Maestro","NVIDIA Nemotron",[42,43,44],"Cadence AI Super Agents","ChipStack","InnoStack",[30,46,47,48],"NVIDIA","Socionext","Silicon Labs",[30,50,51,52,53,40,54,55],"Maestro","EDA","chip design","agentic AI","reinforcement learning","on-premises",1793718868802]