A brand new supercomputer supplied by Cadence, a number one supplier of know-how for digital design automation, is poised to help a collection of engineering design and life sciences purposes accelerated by NVIDIA Blackwell methods and NVIDIA CUDA-X software program libraries.
Accessible to deploy within the cloud and on premises, the Millennium M2000 Supercomputer options NVIDIA HGX B200 methods and NVIDIA RTX PRO 6000 Blackwell Server Version GPUs. Mixed with optimized software program, the supercomputer delivers as much as 80x larger efficiency for digital design automation, system design and life sciences workloads in comparison with its predecessor, a CPU-based system.
With this enhance in computational functionality, engineers can run huge simulations to drive breakthroughs within the design and improvement of autonomous machines, drug molecules, semiconductors, information facilities and extra.
Anirudh Devgan, president and CEO of Cadence, mentioned the collaboration with NVIDIA founder and CEO Jensen Huang onstage at CadenceLIVE, happening as we speak in Santa Clara, California.
“That is years within the making,” Devgan mentioned through the dialog with Huang. “It’s a mixture of development on the {hardware} and system facet by NVIDIA — after which, after all, we have now to rewrite our software program to benefit from that.”
The pair mentioned how NVIDIA and Cadence are working collectively on AI factories, digital twins and agentic AI.
“The work that we’re doing collectively acknowledges that there’s a complete new kind of manufacturing facility that’s needed. We name them AI factories,” Huang mentioned. “AI goes to infuse into each single side of all the pieces we do. Each firm shall be run higher due to AI, or they’ll construct higher merchandise due to AI.”
Huang additionally introduced that NVIDIA plans to buy 10 Millennium Supercomputer methods primarily based on the NVIDIA GB200 NVL72 platform to speed up the corporate’s chip design workflows.
“It is a large deal for us,” he mentioned. “We began constructing our information middle to prepare for it.”
Enabling Clever Design Throughout Industries
The Millennium Supercomputer harnesses accelerated software program from NVIDIA and Cadence for purposes together with circuit simulation, computational fluid dynamics, information middle design and molecular design.
With the supercomputer’s optimized {hardware} and AI software program, engineers and researchers can construct extra advanced, detailed simulations which are able to delivering extra correct insights to allow quicker silicon, methods and drug improvement.
By means of this collaboration, Cadence and NVIDIA are fixing key design challenges with numerous purposes throughout industries — for instance, simulating thermal dynamics for chip design, fluid dynamics for aerospace purposes and molecular dynamics for pharmaceutical analysis.
NVIDIA engineering groups used Cadence Palladium emulation platforms and Protium prototyping platforms to help design verification and chip bring-up workflows for the event of NVIDIA Blackwell.
Cadence used NVIDIA Grace Blackwell-accelerated methods to calculate the fluid dynamics at work when an plane takes off and lands. Utilizing NVIDIA GB200 Grace Blackwell Superchips and the Cadence Constancy CFD Platform, Cadence was capable of run in underneath 24 hours extremely advanced simulations that will take a number of days to finish on a CPU cluster with lots of of hundreds of cores.
Cadence additionally used NVIDIA Omniverse software programming interfaces to visualise these intricate fluid dynamics.
The corporate has built-in NVIDIA BioNeMo NIM microservices into Orion, Cadence’s molecular design platform — and NVIDIA Llama Nemotron reasoning fashions into the Cadence JedAI Platform.
Cadence has additionally adopted the NVIDIA Omniverse Blueprint for AI manufacturing facility digital twins. Related to the Cadence Actuality Digital Twin Platform, the blueprint allows engineering groups to check and optimize energy, cooling and networking in an AI manufacturing facility with bodily primarily based simulations — lengthy earlier than building begins in the true world. With these capabilities, engineers could make quicker configuration choices and future-proof the subsequent technology of AI factories.
Study extra concerning the collaboration between NVIDIA and Cadence and watch this NVIDIA GTC session on advancing physics-based simulation know-how for AI manufacturing facility design.
Photographs courtesy of Cadence.
