28-05-2025
$2M supercomputer powered by NVIDIA delivers 80X leap in chip and drug design
Cadence Design Systems has launched the Millennium M2000 Supercomputer, a powerful new platform built with NVIDIA's latest Blackwell architecture.
Announced during CadenceLIVE Silicon Valley 2025, the system promises significant speed and energy efficiency gains for workloads across semiconductor design, life sciences, aerospace, and hyperscale computing.
Designed for both cloud and on-premises deployment, the Millennium M2000 combines NVIDIA HGX B200 systems, RTX PRO 6000 Blackwell Server Edition GPUs, and Cadence's full stack of solvers.
The system delivers up to 80X faster performance and 20X lower power usage compared to CPU-based predecessors.
By co-optimizing software and hardware, the platform enables simulations that previously required hundreds of CPUs and days to now complete in under 24 hours.
Anirudh Devgan, president and CEO of Cadence, described the system as a milestone for accelerated design. 'The Millennium M2000 Supercomputer will drive the next leap in AI-accelerated engineering by leveraging our massively scalable solvers, dedicated NVIDIA Blackwell-accelerated computing and AI to help designers continue to push the limits of what is possible,' he said.
One key advantage of the M2000 is its ability to handle large-scale simulations in semiconductor and 3D-IC design.
The system supports advanced analysis of power, thermal, stress, and electromagnetic properties in a single unified platform.
According to Cadence, engineers can now complete chip-level power integrity simulations in a single day, compared to the two weeks needed with traditional CPU clusters.
Cadence and NVIDIA engineering teams used the company's Palladium and Protium platforms during the development of Blackwell, helping validate chip designs with faster turnaround times.
'This is years in the making,' Devgan noted. 'It's a combination of advancement on the hardware and system side by NVIDIA — and then, of course, we have to rewrite our software to take advantage of that.'
Beyond chips, the M2000 enables high-fidelity modeling of complex systems, including autonomous machines and digital twin environments.
Designers can now run precision virtual wind tunnels and simulate system behavior at scale, using Cadence tools integrated with NVIDIA's Omniverse APIs. The result is faster development of AI-powered systems in sectors like aerospace and robotics.
NVIDIA founder and CEO Jensen Huang highlighted the growing demand for such infrastructure. 'AI is going to infuse into every single aspect of everything we do. Every company will be run better because of AI, or they'll build better products because of AI,' he said. 'The work that we're doing together recognizes that there's a whole new type of factory that's necessary. We call them AI factories.'
In life sciences, the M2000 supercomputer accelerates molecular design through Cadence's Orion platform, which now integrates NVIDIA BioNeMo NIM microservices and Llama Nemotron models.
Pharmaceutical researchers can explore more design variations faster, helping reduce the time required to identify viable drug candidates.
Cadence said simulations that previously took days using massive CPU clusters can now be completed overnight.
Using the Fidelity CFD Platform and GB200 Grace Blackwell Superchips, the company demonstrated complex fluid dynamics simulations running in under 24 hours.
The Millennium M2000 is expected to cost around $2 million in its standard configuration, which includes approximately 32 NVIDIA Blackwell GPUs.
It is available both via the cloud and as a standalone appliance. Early adopters include MediaTek, Boom Supersonic, Ascendance, Treeline Biosciences, and Supermicro.
NVIDIA also announced plans to acquire 10 Millennium systems to accelerate its own chip design workflows. 'This is a big deal for us,' Huang said. 'We started building our data center to get ready for it.'
