Cadence Unveils NVIDIA-Powered Millennium M2000 Supercomputer

According to reports, Cadence Design Systems has unveiled a powerful new supercomputer built on NVIDIA's latest GPU architecture, aiming to significantly accelerate its software applications across industries ranging from semiconductor design and aerospace engineering to pharmaceutical development.

Cadence, known for providing chip design software to major tech companies such as Apple, has expanded its reach in recent years to serve clients in aviation, biotechnology, and other high-performance sectors. Originally developed for CPUs during the personal computing boom, Cadence has now re-engineered much of its core software to run efficiently on NVIDIA's state-of-the-art Blackwell GPUs.

The newly introduced Millennium M2000 supercomputer is equipped with approximately 32 of NVIDIA's most advanced GPUs. While the most common configuration is expected to cost around $2 million, final pricing details have not yet been disclosed. This marks a significant upgrade from a previous system launched last year, which ran Cadence software with more limited capabilities.

According to Michael Jackson, Vice President and General Manager of the System Design and Analysis Group at Cadence, this performance boost is a game-changer. For example, a simulation task related to turbulence analysis around components of Boeing's 777 aircraft—which previously took eight days to complete on CPU-based systems—can now be executed in less than 24 hours. This allows engineers to not only meet deadlines faster but also to spend additional time refining and optimizing their designs.

"There is an insatiable demand for faster simulation," said Jackson.

At a Cadence event held in Santa Clara, California, NVIDIA CEO Jensen Huang joined Cadence CEO Anirudh Devgan on stage to announce that NVIDIA plans to acquire ten units of the new Millennium M2000 system. These systems will be instrumental in designing the next generation of chips and AI data center infrastructure.

"This is a big deal for us," Huang said. "We've already started building the data centers and preparing for this leap. We aim to increase speed by 50x, 60x, even 100x." NVIDIA will utilize the supercomputers to accelerate its own design and innovation efforts.