Samsung Foundry and Cadence Expand AI Design Ecosystem for Next-Generation AI Infrastructure and Physical AI

Today, AI is driving a fundamental shift in semiconductor design. Delivering next-generation AI systems requires more than advanced process technology—it demands seamless integration across chiplets, 3D-ICs, advanced packaging, high-bandwidth interconnects, memory, and system-level optimization. As the industry moves toward sub-2nm technologies and expands beyond the data center into physical AI, managing system-level complexity has become essential to achieving silicon success.

To address these challenges, Samsung Foundry continues to expand its AI design ecosystem through deep collaboration with leading technology partners. As part of this strategy, Samsung Foundry and Cadence have expanded their multi-year collaboration to deliver a signoff-ready, co-optimized platform built for Samsung Foundry’s 2nm GAA process technology. By combining Samsung Foundry’s advanced process technology with Cadence’s certified AI-driven design flows and broad portfolio of Memory and Interface IP, the collaboration helps customers accelerate the development of next-generation AI infrastructure and physical AI semiconductors.

AI infrastructure is reshaping advanced-node design

Modern AI systems rely on high-bandwidth memory, heterogeneous compute, advanced interconnect fabrics, and multi-die integration, reshaping design priorities at leading-edge process nodes.

Samsung Foundry’s 2nm GAA process is designed to address these evolving requirements by delivering the performance, power efficiency, and scalability needed for AI infrastructure and intelligent edge applications. Samsung is pushing the boundaries of silicon with the new 2nm process, offering superior speed, smaller cell heights, and optimized power efficiency through flexible nanosheet configurations. For high-performance computing (HPC), the performance-optimized 2nm variant introduces hyper cell architecture and low-resistance RDL support to maximize efficiency.

At the same time, process technology is just one of the key factors that enable AI-scale systems. To bring new AI products to market, a tightly integrated design ecosystem spanning implementation, verification, and signoff is critical.

Through its expanded collaboration with Cadence, Samsung Foundry is strengthening this ecosystem with certified AI-driven design flows and an expanded portfolio of advanced Memory and Interface IP, helping customers develop AI systems, chiplets and 3D-IC architectures optimized for power, performance and area (PPA).

Building an integrated platform for AI system design

Delivering next-generation AI systems require close integration across process technology, IP, EDA, advanced packaging, and interconnects to help customers manage increasing design complexity.

Samsung Foundry and Cadence are addressing this challenge through a co-optimized platform that combines Samsung Foundry’s 2nm GAA process technology with Cadence’s certified implementation, verification, analysis, and signoff solutions. The platform also includes an expanded portfolio of Memory and Interface IP, including PCIe 7.0, UCIe 64G, HBM5 and 224G/256G SerDes.

In addition, support for NVIDIA CUDA-X GPU-accelerated libraries enables Cadence’s implementation, analysis, and signoff solutions to run key workloads on Cadence Millennium Supercomputers built on NVIDIA accelerated computing platform. The collaboration also extends to Samsung Foundry’s advanced packaging technologies, including 3D Cube-H, a vertical integration technology using Hybrid Copper Bonding (HCB), enabling tighter integration across implementation, analysis, packaging, and signoff workflows.

Together, these capabilities help customers accelerate design convergence, improve performance, power, and area (PPA), reduce development risk, and bring increasingly complex AI infrastructure, chiplet-based architectures, and physical AI systems to market faster.

Industry leaders are accelerating AI innovation on Samsung Foundry’s 2nm platform

The expanded Samsung Foundry–Cadence collaboration is already helping leading AI innovators accelerate the development of next-generation AI infrastructure and physical AI systems.

Enabling the next-generation of AI infrastructure

NVIDIA is leveraging Cadence’s GPU-accelerated design flows on Samsung Foundry’s second-generation 2nm platform to optimize next-generation accelerated computing architectures and high-bandwidth interconnect technologies.

Advancing physical AI at the intelligent edge

Ambarella is leveraging Samsung Foundry’s advanced 2nm process and Cadence’s implementation, IP, and system analysis technologies to accelerate its next-generation edge AI platform. Designed for robotics, drones, and autonomous machines, the platform supports real-time multimodal AI, scaling VLMs and LLMs at the edge with industry-leading performance-per-watt.

Unlike cloud-scale AI accelerators, physical AI systems must integrate AI compute with multi-sensor processing, sensor fusion, real-time responsiveness, low-power operation, and high reliability within compact edge environments. Meeting these requirements demands highly optimized implementation, verification, and signoff workflows at advanced process nodes.

According to Ambarella COO Chan Lee, Cadence’s PCIe connectivity solutions, together with Samsung Foundry’s advanced 2nm technology and a production-proven design ecosystem, have helped simplify the development of next-generation edge AI devices.

“Ambarella’s edge AI strategy is focused on delivering industry leading performance per watt, scalable AI acceleration, and robust multi sensor processing at the most advanced process nodes. Our collaboration with Cadence and Samsung Foundry to deliver IP for PCIe 5.0 for our next generation 2nm edge AI platform has been critical as we address the design, verification and manufacturing complexity of this node.”

— Chan Lee, COO, Ambarella

Accelerating the future of AI system design

The Samsung Foundry–Cadence collaboration reflects a broader shift in semiconductor development. As AI systems become increasingly complex, leadership at advanced nodes will be defined not only by transistor innovation, but also by the strength of the ecosystem that enables customers to move from architecture to silicon with greater speed and confidence.

By combining advanced process technology, AI-driven design methodologies, and a robust partner ecosystem, Samsung Foundry and Cadence are helping customers reduce design complexity and accelerate the next generation of AI infrastructure and physical AI systems.

As the semiconductor industry enters the era of infrastructure AI and physical AI, ecosystem collaboration will continue to play a critical role in transforming silicon innovation into real-world intelligence.