{"id":3749,"date":"2026-03-20T19:05:03","date_gmt":"2026-03-20T23:05:03","guid":{"rendered":"https:\/\/blogs.sw.siemens.com\/calibre\/?p=3749"},"modified":"2026-03-26T16:24:38","modified_gmt":"2026-03-26T20:24:38","slug":"from-data-to-domain-expertise-how-ai-accelerated-computing-and-digital-twins-are-reshaping-semiconductor-manufacturing","status":"publish","type":"post","link":"https:\/\/blogs.sw.siemens.com\/calibre\/2026\/03\/20\/from-data-to-domain-expertise-how-ai-accelerated-computing-and-digital-twins-are-reshaping-semiconductor-manufacturing\/","title":{"rendered":"From data to domain expertise: How AI, accelerated computing and digital twins are reshaping semiconductor manufacturing"},"content":{"rendered":"\n<p><em>The dawn of industrial AI in semiconductor manufacturing<\/em><\/p>\n\n\n\n<p>A transformative moment is unfolding in the semiconductor manufacturing industry\u2014a shift driven by artificial intelligence (AI), digital twins and GPU-accelerated computing. At the 2026 NVIDIA GTC, <a href=\"https:\/\/www.linkedin.com\/in\/danping-peng-3848942\/\" target=\"_blank\" rel=\"noreferrer noopener\">Danping Peng<\/a>, vice president of research and development for the Calibre semiconductor manufacturing group at Siemens EDA, shared his insights into how these advancements are setting the foundation for a smarter, more agile industrial future.<\/p>\n\n\n\n<figure class=\"wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe loading=\"lazy\" title=\"NVIDIA GTC 2026 Live\" width=\"640\" height=\"360\" src=\"https:\/\/www.youtube.com\/embed\/RTmSrIFZanc?start=2032&#038;feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure>\n\n\n\n<blockquote class=\"wp-block-quote is-layout-flow wp-block-quote-is-layout-flow\">\n<p>As Peng explained, \u201cWe\u2019re entering an era where AI is not just assisting our experts\u2014it\u2019s empowering every engineer to harness data, collaborate seamlessly and make better decisions instantly.\u201d<\/p>\n\n\n\n<p><\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Building value from data: The foundation for smarter manufacturing<\/h2>\n<\/blockquote>\n\n\n\n<p>Every device, vehicle and industrial process today generates massive streams of data. Data in the information age serves as the \u201coil\u201d\u2014a core asset packed with potential. But extracting actionable information from this abundance requires algorithms that are both powerful and purpose-built. Peng said that turning data into value starts with domain context. In semiconductor manufacturing, it\u2019s essential that our models understand what matters to engineers and operators\u2014accuracy, speed and safety.<\/p>\n\n\n\n<p>Large language models represent a leap in AI versatility, but as Peng explained, \u201cbreadth comes with weakness.\u201d These models are often generalists (\u201cJack of every trade, master of none\u201d). For semiconductor professionals striving for precision and efficiency, the next step is clear: agentic AI.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Agentic AI: Turning domain knowledge into automation<\/h2>\n\n\n\n<p>Agentic AI brings practical, domain-specific intelligence to the world of semiconductor manufacturing. Rather than relying on AI models trained only on the world\u2019s data, agentic AI transforms the know-how of expert engineers into digital agents. These agents use natural language, orchestrate complex workflows and deliver outcomes\u2014often making specialized tasks accessible to a broader audience, including non-technical stakeholders.<\/p>\n\n\n\n<p>As Peng summarized, \u201cAgentic AI allows us to capture the wisdom of our best engineers and scale it across the organization. Now, teams can automate multi-step processes that used to require years of specialized experience.\u201d<\/p>\n\n\n\n<p>Practically, this means that planning, acting and delivering results no longer require deep programming knowledge. Teams can activate high-value automation, accelerating productivity and focusing human expertise where it matters most. For investors and stakeholders, the result is improved ROI driven by faster development and smarter decision-making.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"972\" height=\"1024\" src=\"https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/agentic_ai-972x1024.png\" alt=\"A graphical depiction of the layered architecture of industrial AI systems. At the top, three application types\u2014industrial copilots (robot icon), software applications (computer icon), and industrial machines (robot arm icon)\u2014interface with users. Below these, orchestration manages agent workflows. The agents layer shows dedicated industrial capabilities (various icons), which connect to the foundation models layer at the bottom, representing models trained with industrial data. Lines show how foundation models support agents, which are orchestrated to serve applications. \" class=\"wp-image-3750\" srcset=\"https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/agentic_ai-972x1024.png 972w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/agentic_ai-569x600.png 569w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/agentic_ai-768x809.png 768w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/agentic_ai-900x948.png 900w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/agentic_ai.png 1025w\" sizes=\"auto, (max-width: 972px) 100vw, 972px\" \/><\/figure>\n\n\n\n<p><a href=\"https:\/\/resources.sw.siemens.com\/en-US\/technical-paper-al-and-ml-in-calibre-ic-manufacturing-the-intelligent-solutions-driving\/\" target=\"_blank\" rel=\"noreferrer noopener\">Read more how Calibre manufacturing tools use AI\/ML.<\/a> <\/p>\n\n\n\n<h2 class=\"wp-block-heading\">The Siemens EDA and NVIDIA partnership<\/h2>\n\n\n\n<p>Earlier this year, Siemens EDA <a href=\"https:\/\/press.siemens.com\/global\/en\/pressrelease\/siemens-and-nvidia-expand-partnership-build-industrial-ai-operating-system\" target=\"_blank\" rel=\"noreferrer noopener\">announced<\/a> an expanded partnership with NVIDIA, targeting four core areas:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>AI-driven simulation<\/li>\n\n\n\n<li>Supply chain management<\/li>\n\n\n\n<li>Manufacturing optimization<\/li>\n\n\n\n<li>Electronic design automation (EDA)<\/li>\n<\/ul>\n\n\n\n<p>Siemens EDA sits at the nexus of manufacturing, engineering and digital transformation. Through our Xcelerator platform\u2014a portfolio of software, hardware and services built to accelerate industrial innovation\u2014the company supports global manufacturers in designing products, managing supply chains and optimizing production.<\/p>\n\n\n\n<p>Peng emphasized the significance of this partnership: \u201cOur collaboration with NVIDIA means simulation and automation can happen faster than ever before, with true-to-life context and physics. We\u2019re leveraging industry-leading GPUs to drive the next decade of productivity for chip and system makers.\u201d<\/p>\n\n\n\n<p>By leveraging NVIDIA\u2019s Omniverse, enabled Xcelerator with real-time collaboration among distributed engineering teams in photorealistic digital environments. The result: digital twins of manufacturing processes that provide instant feedback on the impact of process changes\u2014a decisive advantage in time-to-value and operational agility.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"577\" src=\"https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/Siemens-and-Nvidia-1024x577.jpg\" alt=\"Jensen Huang (on the right), founder and CEO of NVIDIA, and Roland Busch (on the left), President and CEO of Siemens AG, shaking hands on stage at CES 2026.\" class=\"wp-image-3751\" srcset=\"https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/Siemens-and-Nvidia-1024x577.jpg 1024w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/Siemens-and-Nvidia-600x338.jpg 600w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/Siemens-and-Nvidia-768x433.jpg 768w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/Siemens-and-Nvidia-1536x866.jpg 1536w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/Siemens-and-Nvidia-2048x1154.jpg 2048w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/Siemens-and-Nvidia-395x222.jpg 395w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/Siemens-and-Nvidia-900x507.jpg 900w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\">Digital twins at scale: Simulating the factory of the future<\/h2>\n\n\n\n<p>The concept of the digital twin\u2014a comprehensive, dynamic virtual representation of assets, products and processes\u2014stands at the heart of digital transformation. With Siemens Xcelerator and Omniverse integration, digital twins now scale from singular processes to entire factory floors, vehicles and infrastructure ecosystems.<\/p>\n\n\n\n<p>Peng explained, \u201cThe digital twin transforms how we design, build, and run manufacturing environments. With real-time simulation, engineering becomes more predictive and less reactive.\u201d<\/p>\n\n\n\n<p>These high-fidelity simulations allow manufacturers to test scenarios, analyze outcomes and validate designs before any physical investment. Feedback loops between the virtual and physical worlds become nearly instantaneous. Operational risk drops, while opportunities for optimization skyrocket\u2014a fundamental enabler for next-generation, resilient supply chains.<\/p>\n\n\n\n<figure class=\"wp-block-image size-large\"><img loading=\"lazy\" decoding=\"async\" width=\"1024\" height=\"576\" src=\"https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/digital-twin-1024x576.png\" alt=\"Split-screen image illustrating semiconductor technology and industrial automation. The left side features a dark background with a light blue outline of a microchip, resembling a digital schematic. The right side shows a robotic arm holding an actual microchip, representing advanced manufacturing and precision robotics in a modern factory setting.\" class=\"wp-image-3752\" srcset=\"https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/digital-twin-1024x576.png 1024w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/digital-twin-600x338.png 600w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/digital-twin-768x432.png 768w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/digital-twin-1536x864.png 1536w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/digital-twin-2048x1152.png 2048w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/digital-twin-395x222.png 395w, https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/digital-twin-900x506.png 900w\" sizes=\"auto, (max-width: 1024px) 100vw, 1024px\" \/><\/figure>\n\n\n\n<h2 class=\"wp-block-heading\">Physics AI: Embedding the laws of the physical world into machine learning<\/h2>\n\n\n\n<p>Traditional AI models in manufacturing rely heavily on large volumes of historical and simulated data. However, collecting and annotating such data remains expensive and time-consuming, especially as semiconductor geometries shrink and manufacturing complexity increases.<\/p>\n\n\n\n<p>Peng highlighted the opportunity: \u201cPhysics-based foundation models help us bridge the gap between what our engineers know and what our data can tell us. By embedding the laws of physics, we get the best of both worlds: trustworthy AI and faster results.\u201d<\/p>\n\n\n\n<p>Physics AI offers a way forward. By embedding physics-based constraints directly into AI models, Siemens EDA builds solutions that are more stable, predictive and faster to train. This approach narrows the gap between simulated results and real-world behavior, pushing beyond the limits of purely data-driven methods.<\/p>\n\n\n\n<p>As sensor data, process telemetry and operational feedback are fused with computational models, manufacturers get systems capable of near real-time self-correction and optimization. The promise: robust, lower-cost manufacturing and rapid innovation cycles\u2014even as traditional big data approaches hit scalability barriers.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Accelerated computing: The GPU-powered future of EDA and production<\/h2>\n\n\n\n<p>The complexity of semiconductor manufacturing and design today is staggering. Modern GPUs, for example, vastly outstrip their predecessors in raw parallelism and system integration, demanding new approaches to design, verification and yield management.<\/p>\n\n\n\n<p>\u201cGPUs have changed the game,\u201d noted Peng. \u201cThey enable us to simulate, validate, and orchestrate designs on a scale we could only dream of a few years ago.\u201d<\/p>\n\n\n\n<p>GPU acceleration is now foundational in many Siemens EDA tools. It powers not only the visualization and manipulation of digital twins but also the solution of intricate physical equations at industrial scale. This shift enables orders-of-magnitude reduction in simulation and verification times, unlocking previously impossible optimization routines and design space exploration.<\/p>\n\n\n\n<p>NVIDIA\u2019s CUDA platform further extends these gains, empowering AI models that proactively monitor and correct manufacturing workflows in real time. For investors and engineering leaders, the message is clear: the tools shaping the next decade of silicon are being born on the platforms of today.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">Embedded intelligence: Bringing generative and agentic AI to every workflow<\/h2>\n\n\n\n<p>In semiconductor manufacturing and EDA, automation is only as powerful as its ability to reason, adapt and act. The Siemens EDA <a href=\"https:\/\/www.siemens.com\/en-us\/products\/fuse-eda-ai-system\/\" target=\"_blank\" rel=\"noreferrer noopener\">Fuse EDA AI system<\/a> is engineered to embed generative and agentic AI directly into every major workflow\u2014from initial circuit design through manufacturing of integrated circuits and printed circuit board (PCB) systems.<\/p>\n\n\n\n<p>\u201cOur vision,\u201d Peng said, \u201cis to put reasoning AI right inside the loop for every engineer, designer, and operator. By blending agentic and generative models, we make innovation accessible\u2014moving from complicated software to collaborative intelligence.\u201d<\/p>\n\n\n\n<p>This cohesive AI strategy ensures that knowledge is shared, processes are streamlined and complexity does not stand in the way of innovation. With the expanded NVIDIA partnership announced at GTC, Siemens EDA now offers AI-powered reasoning and autonomous process management in the IC and PCB product lines\u2014changing the nature of engineering collaboration.<\/p>\n\n\n\n<p>The implications reach beyond speed. AI agents can now automate multi-step workflows, resolve bottlenecks and perform tradeoff analyses that once demanded expert intervention, freeing designers to focus on creativity and differentiation.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">The future: Smarter factories, faster operations and greater value<\/h2>\n\n\n\n<p>Semiconductor professionals, investors and OEMs are witnessing an unprecedented transformation. With its deep domain expertise and ecosystem partnerships, Siemens Digital Industries Software is positioned to enable a future where industrial teams collaborate with AI agents, digital twins bridge the virtual and real and the boundaries of what\u2019s possible are continually redefined.<\/p>\n\n\n\n<p>By harnessing agentic AI, physics-based modeling and accelerated computing, organizations can expect:<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li>Dramatically improved time-to-market and time-to-productivity<\/li>\n\n\n\n<li>Resilient, flexible and efficient supply chains<\/li>\n\n\n\n<li>Lower operational costs and fewer process failures<\/li>\n\n\n\n<li>Higher-quality, more affordable products for every market segment<\/li>\n<\/ul>\n\n\n\n<p>Peng summarized, \u201cThis is an inflection point. The integration of AI, digital twins, and GPU acceleration will unlock the next wave of productivity, resilience, and creativity for the world\u2019s most demanding industries.\u201d<\/p>\n\n\n\n<h4 class=\"wp-block-heading\">Join the conversation: Watch the recorded NVIDIA GTC livestream<br><a href=\"https:\/\/lnkd.in\/g7CUP6us\" target=\"_blank\" rel=\"noopener\">Watch on LinkedIn<\/a> or <a href=\"https:\/\/lnkd.in\/g-Kki9jP\" target=\"_blank\" rel=\"noopener\">on YouTube<\/a><\/h4>\n","protected":false},"excerpt":{"rendered":"<p>Explore the dawn of industrial AI in semiconductor manufacturing. Learn how Siemens EDA is leveraging AI, digital twins, and GPU-accelerated computing to empower engineers, automate workflows, and build smarter factories.<\/p>\n","protected":false},"author":71645,"featured_media":3752,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"spanish_translation":"","french_translation":"","german_translation":"","italian_translation":"","polish_translation":"","japanese_translation":"","chinese_translation":"","footnotes":""},"categories":[1],"tags":[963,958,22,964],"industry":[],"product":[90],"coauthors":[712],"class_list":["post-3749","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","tag-accelerated-computing","tag-ai","tag-digital-twin","tag-semiconductor-manufacturing","product-calibre"],"featured_image_url":"https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/50\/2026\/03\/digital-twin.png","_links":{"self":[{"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/posts\/3749","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/users\/71645"}],"replies":[{"embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/comments?post=3749"}],"version-history":[{"count":3,"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/posts\/3749\/revisions"}],"predecessor-version":[{"id":3756,"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/posts\/3749\/revisions\/3756"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/media\/3752"}],"wp:attachment":[{"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/media?parent=3749"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/categories?post=3749"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/tags?post=3749"},{"taxonomy":"industry","embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/industry?post=3749"},{"taxonomy":"product","embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/product?post=3749"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/calibre\/wp-json\/wp\/v2\/coauthors?post=3749"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}