케이던스 디자인 시스템즈(Cadence Design Systems, Inc.)는 5일 업계 최초로 디지털 트윈의 다중물리(Multi-physics) 시스템 설계 및 분석을 가속화할 수 있는 HW/SW 솔루션 ‘케이던스 밀레니엄 엔터프라이즈 다중물리 플랫폼(Cadence Millennium Enterprise Multiphysics Platform)’을 발표했다.
Millennium Platform with Built-in GPU, Up to 100x Design Efficiency
Cadence's new Cadence Millennium Platform accelerates and scales digital twin and AI applications.
Cadence Design Systems, Inc. announced on the 5th the 'Cadence Millennium Enterprise Multiphysics Platform', a HW/SW solution that can accelerate the design and analysis of multi-physics systems of digital twins for the first time in the industry.
The first-generation Cadence Millennium M1 accelerates high-performance Computational Fluid Dynamics (CFD) simulations. Turnkey Cadence high-fidelity CFD software with optimized GPU acceleration and fast interconnects for generative AI is included.
Millennium M1 provides customers with unprecedented turnaround times and near-linear scalability when simulating complex mechanical systems by fusing multiple stages into an integrated cluster.
Designing new levels of performance and efficiency in mechanical systems has become a key priority for the automotive, aerospace and defense, energy and turbomachinery industries. “Advances in multi-physics simulation technology are essential to achieving these goals,” Cadence said. “Accelerating speed, accuracy, capacity and compute speed are essential elements of digital twin simulations, enabling design innovations that can provide confidence that a design will perform as intended before prototypes are developed and tested.”
The Cadence Millennium Platform combines a best-in-class GPU-embedded CFD solver with dedicated GPU hardware, delivering supercomputer-class throughput per GPU with up to 1,000 CPU cores. It also delivers up to 20x better energy efficiency than comparable CPUs, reducing processing times from weeks to hours.
It utilizes the 'Fidelity CFD Solver' which can solve complex simulation problems and provides excellent accuracy. It provides near-linear scalability across multiple GPU nodes by building a scalable architecture and a highly flexible fidelity solver.
By rapidly generating high-quality multi-physics data, generative AI can quickly and reliably create digital twin visualizations of optimal system design solutions.
“For 35 years, Cadence has focused on improving computational speed while maintaining accuracy in very difficult computing areas,” said Ben Gu, vice president of R&D for Multi-physics System Analysis at Cadence. “Algorithmic throughput remains a key priority, and we are now leveraging generative AI to extract insights from massive amounts of design and simulation data.”
“The groundbreaking Millennium platform is a giant leap forward in accelerating and scalable digital twin and AI applications, and as CFD becomes more powerful and efficient, the exceptional performance of the Millennium M1 will revolutionize industries that require rapid time to market,” he said.