CadenceLIVE 2025 – Driving AI Inflection Point in Design and Simulation Software
0
May 12, 2025
At CadenceLIVE 2025, the company announced its latest Agentic AI-based software platform enabling automation design and implementation workflows.
Cadence also announced the Millenium M2000 Supercomputer accelerated by NVIDIA’s Blackwell platform to boost designer productivity, advancing the development of next-generation AI infrastructure, physical AI systems, and drug discovery.
Cadence-NVIDIA’s partnership is one of the most symbiotic where NVIDIA leverages Cadence’s software platform to develop advanced semiconductor chips while Cadence uses a NVIDIA-based supercomputer to help companies accelerate design run times and time-to-market.
Cadence, a leading semiconductor and electronic system design, simulation and implementation software, tool and IP company, recently concluded its annual CadenceLIVE 2025 event held at Silicon Valley, USA. The company announced a series of innovations and partnerships driving the inflection point of AI integration in its design and simulation software platform and tools.
Some of the key announcements were:
Cadence Cerebrus® AI Studio: The company announced Cerebrus AI Studio, claimed to be the industry’s first multi-block, multi-user “automated” design platform powered by Agentic AI. The AI-driven platform can take multi-step decisions and actions for block-level orchestration, optimization, and implementation of SoC design workflows.
Cadence aims to accelerate the entire SoC design to development cycle thereby boosting time-to-market by almost 5x. This is achieved by leveraging Agentic AI-driven design workflows. The Agentic-AI framework also helps in achieving significant gains in PPA (~20% more) for designing complex semiconductors with multi-design, multi-run approach, advanced analytics and accumulated learning to optimize the designs quickly.
Source: Cadence
Key Takeaways:
As we enter the AI era, the AI accelerator chip is approaching almost angstrom-level process nodes, packing billions of transistors and multiple compute and memory blocks with a more complex 2.5D to 3D architecture, from transistors to packaging.
These specialized chips create significant design complexity and are challenged by accelerated design cycles and limited design talent.
So, an agentic AI-driven automated approach to design can solve most of the above issues with multi-step, hierarchical, intelligent workflows driving both performance and productivity (~10x).
The multi-agent approach for various stages of the design to implementation lifecycle should assist design engineers to cut down manual tasks, thus shortening the design cycle as well as boosting accuracy.
Candence’s Millenium M2000 Supercomputer and NVIDIA Partnership:
Cadence has further enhanced the integration of its advanced design and simulation software, Millennium Enterprise Platform, with the introduction of the new Millennium M2000 Supercomputer. This system is accelerated by NVIDIA’s Blackwell HGX B200, RTX PRO 6000 Blackwell Server Edition GPUs, and NVIDIA’s CUDA-X libraries. Specifically designed for innovative AI models, the Millennium M2000 Supercomputer significantly boosts designer productivity, advancing the development of next-generation AI infrastructure, physical AI systems, and drug discovery. It aims to do so by reducing simulation run times and boost EDA, SDA, and other complex workloads performance by almost 80x and power efficiency by 20x vs CPU-based processing of design and simulation workloads.
Source: Cadence
Key Takeaways:
The Move from CPUs to GPU-accelerated Simulation Design:
Enabling complex simulations of fluid dynamics for the aerospace industry.
Source: NVIDIA
Firstly, the industry is moving from two-dimensional to a three-dimensional representation of the world. This is seen from basic semiconductor design of 2D to 3D silicon architectures for both compute and memory to spatial intelligence with AR/VR and autonomous vehicles to robotics.
For designing those advanced 3D chipsets, 3D molecules or even LiDAR, Camera and sensor-based autonomous mobility and spatial movements for cars, aerospace, more complex synthetic data generation is required to train the models for advanced simulations.
Secondly, to process this type of advanced use-cases, AI is critical and hence a move from Generative AI to Physical AI is necessary. GPUs are the only compute accelerators that can process these complex workloads at scale and efficiency. Hence, by shifting to GPUs, Cadence is equipping enterprises with the essential power to navigate this 3D, AI-accelerated future.
Benefits to Cadence’s Customers:
This is a significant move by Cadence as the company is offering a service to enterprise customers and startups joining the 3D, AI-accelerated future with no resources to build their own GPU-scale infrastructure as it is a high CAPEX investment.
Enterprises that have 3D-simulated design at the center of their R&D and product development will likely see this offering from Cadence (on cloud or on-prem) reduce a big barrier in terms of time-to-market while helping them scale their offerings.
This democratization of access to advanced software will drive innovations and enterprises forward, allowing enterprises to invest and focus on their core skills and leverage AI for success as they grow instead of being sidetracked building their own huge expensive IT infrastructure.
Cadence-NVIDIA Partnership – A Symbiotic One:
Cadence has incorporated NVIDIA’s Blackwell hardware and CUDA-X software into its systems. It integrated NVIDIA’s BioNeMo NIM microservices with Orion, the molecular design platform, and NVIDIA’s Llama Nemotron models with the JedAI Platform. Additionally, Cadence uses the NVIDIA Omniverse Blueprint with the Reality Digital Twin Platform to help engineers test and optimize power, cooling, and networking for AI factories using simulations before construction begins. Cadence also used NVIDIA’s Omniverse APIs to visualize these intricate fluid dynamics, molecular drug, or datacenter design simulations.
NVIDIA's design team used Cadence Palladium emulation and Protium prototyping platforms for design verification and chip bring-up for the development of Blackwell and future chip platforms.
Sources: NVIDIA, Cadence
Key Takeaways:
NVIDIA and Cadence’s collaboration is also a notable example of a ‘symbiotic interdependent’ relationship.
To design advanced accelerators, NVIDIA needs Cadence's software and tools. On the other hand, to help enterprises such as NVIDIA design advanced semiconductors, Cadence needs to have software and AI capabilities optimized on the greatest accelerator infrastructure which can make this 3D design and simulation possible.
Hence, NVIDIA's decision to also purchase M2000, sporting its own solutions, speaks volumes on how the entire AI ecosystem works in a collaborative and highly integrated manner.
Sometimes horizontal integration is required and sometimes vertical integration in more ideal to allow companies help themselves and their customers scale efficiently.
Neil is a sought-after frequently-quoted Industry Analyst with a wide spectrum of rich multifunctional experience. He is a knowledgeable, adept, and accomplished strategist. In the last 18 years he has offered expert strategic advice that has been highly regarded across different industries especially in telecom. Prior to Counterpoint, Neil worked at Strategy Analytics as a Senior Analyst (Telecom). Neil also had an opportunity to work with Philips Electronics in multiple roles. He is also an IEEE Certified Wireless Professional with a Master of Science (Telecommunications & Business) from the University of Maryland, College Park, USA.
Related Research
May 8, 2025
Apr 4, 2025
Mar 20, 2025
Jan 15, 2025