Investment Insights - A Clever Defense by NVIDIA

The two giants of the semiconductor ecosystem, NVIDIA and Synopsys, have announced a milestone multi-year strategic collaboration, with NVIDIA investing $2 billion in a project led by Synopsys. This collaboration aims to integrate NVIDIA's GPU-accelerated computing platform with Synopsys' industry-leading electronic design automation (EDA) and semiconductor IP product portfolio, significantly accelerating chip design cycles, greatly reducing power consumption, and supporting the development of next-generation artificial intelligence, automotive, and high-performance computing chips.

The core of this collaboration is to create a unified cloud-native design environment that integrates Synopsys' TestMAX, Verdi, and VC Formal tools, NVIDIA's cuLitho computational lithography platform, and the broader Grace-Blackwell software stack. Designers will be able to run full-chip layout, design rule checks, and electromagnetic simulations at GPU-accelerated speeds for the first time, achieving speeds 10 to 50 times faster than traditional CPU-based processes. Early benchmarks shared at the launch event indicated that the same sign-off process for a 3nm AI training chip, which previously required 12 weeks of computation time, could be completed in less than 60 hours on NVIDIA DGX cloud instances running Synopsys tools.

NVIDIA CEO Jensen Huang described this collaboration as "the most significant turning point in the EDA field since the invention of logic synthesis technology." He emphasized that as the number of transistors on chips approaches trillions and enters the exascale era, traditional Moore's Law scaling is being replaced by system-level innovations that "go beyond Moore's Law." "The bottleneck is no longer just the foundry," Huang said at the Synopsys SNUG conference in Santa Clara, "but the months designers spend waiting for verification regression test results. We are working together to eliminate this bottleneck."

Synopsys CEO Sassine Ghazi revealed that the $2 billion investment (combining direct R&D funding, pre-purchased cloud credits, and joint engineering personnel) will be used to drive three flagship initiatives:

Synopsys has launched a new AI-driven EDA suite called "Synopsys.ai Copilot," built on NVIDIA's BlueField-3 DPU and Grace CPU, which can suggest optimal layout solutions, predict timing convergence, and automatically generate test platforms using large graph models trained on decades of Synopsys design data.

Fully integrating NVIDIA's cuPPA (CUDA Power and Performance Analysis) toolkit into Synopsys PrimePower to achieve Peta-level accurate dynamic power simulation across multi-chip systems, which is crucial for next-generation AI accelerators and autonomous vehicle SoCsAn open "NVIDIA-Synopsys Foundry Design Kit" program will provide pre-validated reference flows for TSMC's 2nm and Intel's 18A process nodes, significantly lowering the barrier for startups and hyperscale companies to tape out complex chip designs.

Analysts praised the deal as a clever defense by NVIDIA, whose dominance in AI training hardware is facing increasing scrutiny. By locking in Synopsys—an EDA market leader with over 55% market share in digital design—NVIDIA effectively strengthens the competitive barriers of its ecosystem. Competitors (such as Cadence/Siemens EDA) now face the risk that their flagship tools may not perform optimally without NVIDIA chips.

Perhaps most notably, sources close to both companies revealed that the collaboration agreement includes a low-key but far-reaching clause: any chips designed using the jointly developed process must include a "design watermark" created by NVIDIA in the GDSII files—an encrypted signature that can be read by NVIDIA's cloud orchestration layer. Although both companies insist that this move is purely for performance benchmarking and royalty-free intellectual property tracking, critics have labeled it a "tax on the future of fabless companies."

However, for frontline engineers, the prospects brought by this technology are both simple and profound: chip tape-out work that previously took nine months could soon be reduced to nine weeks. As one SoC architect based on Arm architecture told me after a keynote speech, "If this technology succeeds, we can not only design chips faster but also design chips that were previously impossible to create."

With companies like Samsung, Broadcom, and MediaTek signing on as early customers, the collaboration between NVIDIA and Synopsys is reshaping the landscape of semiconductor innovation. The era of GPU-accelerated chip design has officially begun.

NVIDIA and Synopsys Announce Strategic Partnership to Revolutionize Engineering and Design

NVIDIA and Synopsys recently announced an expansion of their strategic partnership to fundamentally change design and engineering across various industries.

From the semiconductor industry to aerospace, automotive, and industrial sectors, R&D teams face numerous engineering challenges, including increasingly complex workflows, rising development costs, and pressure to shorten time-to-market. This expanded partnership will integrate NVIDIA's strengths in artificial intelligence and accelerated computing with Synopsys' market-leading engineering solutions, helping R&D teams design, simulate, and verify intelligent products with greater accuracy, speed, and lower costs. Additionally, NVIDIA invested $2 billion to purchase common stock of Synopsys at a price of $414.79 per share.

"CUDA GPU-accelerated computing is revolutionizing design—it enables simulation at unprecedented speed and scale, creating fully functional digital twin models from atoms to transistors, from chips to complete systems," said Jensen Huang, founder and CEO of NVIDIA"Our collaboration with Synopsys fully leverages the powerful capabilities of NVIDIA accelerated computing and artificial intelligence to reimagine engineering and design—empowering engineers to create extraordinary products that shape the future."

"The complexity and cost of developing next-generation intelligent systems require engineering solutions that integrate electronics and physics more deeply, while accelerating this process with artificial intelligence and computing power. No two companies are better positioned than Synopsys and NVIDIA to provide AI-driven holistic system design solutions," said Sassine Ghazi, President and CEO of Synopsys. "Together, we will reshape engineering, empowering innovators around the world to achieve innovation more efficiently."

This multi-year collaboration builds on a strong existing technological partnership between the two companies and includes the following initiatives:

Comprehensive acceleration of Synopsys applications: Utilizing NVIDIA CUDA-X libraries and AI-Physics technology, Synopsys will further accelerate and optimize its extensive portfolio of compute-intensive applications, covering areas such as chip design, physical verification, molecular simulation, electromagnetic analysis, and optical simulation.

Advancing intelligent AI engineering: Building on existing AI collaborations, the two companies are integrating Synopsys AgentEngineer™ technology with NVIDIA's intelligent AI technology stack (including NVIDIA NIM microservices, NVIDIA NeMo Agent Toolkit software, and NVIDIA Nemotron models) to achieve autonomous design capabilities for EDA, simulation, and analysis workflows.

Connecting the physical and digital worlds through digital twins: The companies will collaborate to leverage high-precision, high-complexity digital twin technology to create next-generation virtual design, testing, and verification solutions for industries such as semiconductors, robotics, aerospace, automotive, energy, industrial, and healthcare. These solutions will utilize NVIDIA Omniverse, NVIDIA Cosmos, and other technologies.

Cloud-ready solutions: Synopsys and NVIDIA plan to enable cloud access for GPU-accelerated engineering solutions, allowing engineering teams of all sizes to benefit from the powerful capabilities of accelerated engineering solutions.

Developing a joint marketing plan: To drive market acceptance, both parties have also agreed to develop a joint marketing plan to broadly cover engineering teams across multiple industries through on-premises and cloud solutions. This marketing effort will leverage Synopsys's extensive network of thousands of direct sales and channel partners worldwide, supported by Synopsys's large customer base and the existing Omniverse technology licensing, sales, and support agreements between the two companies (which have been embedded in Synopsys's simulation solutions) This collaboration is not exclusive. NVIDIA and Synopsys will continue to work with the broader semiconductor and electronic design automation (EDA) ecosystem to create opportunities for mutual growth in the fields of engineering and design.

Risk Warning and Disclaimer

The market has risks, and investment requires caution. This article does not constitute personal investment advice and does not take into account the specific investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. Investment based on this is at one's own risk