In a groundbreaking improvement, NVIDIA Omniverse and Modulus are empowering Ansys to speed up 3D simulation workflows, a key element in constructing next-generation semiconductor methods, in keeping with the NVIDIA Weblog.
The Function of 3D-ICs in Semiconductor Design
Multi-die chips, often known as three-dimensional built-in circuits (3D-ICs), signify a revolutionary step in semiconductor design. These chips are vertically stacked to create a compact construction that enhances efficiency with out rising energy consumption. Nonetheless, as chips change into denser, they current extra complicated challenges in managing electromagnetic and thermal stresses. Superior 3D multiphysics visualizations change into important to design and diagnostic processes to handle these challenges.
Insights from the Design Automation Convention
On the Design Automation Convention, a worldwide occasion showcasing the most recent developments in chips and methods, Ansys demonstrated the way it leverages NVIDIA expertise to beat these challenges. Ansys makes use of NVIDIA Omniverse, a platform of software programming interfaces, software program improvement kits, and providers, to allow 3D visualizations of simulation outcomes. This platform powers visualizations of 3D-IC outcomes from Ansys solvers, permitting engineers to judge phenomena like electromagnetic fields and temperature variations to optimize chips for quicker processing, elevated performance, and improved reliability.
Enhanced Simulation Capabilities
With Ansys Icepak on the NVIDIA Omniverse platform, engineers can simulate temperatures throughout a chip in keeping with completely different energy profiles and flooring plans. Figuring out chip hot-spots can result in higher chip designs and auxiliary cooling units. Nonetheless, these 3D-IC simulations are computationally intensive, limiting the variety of simulations and design factors customers can discover.
Utilizing NVIDIA Modulus, mixed with novel strategies for dealing with arbitrary energy patterns within the Ansys RedHawk-SC electrothermal knowledge pipeline and mannequin coaching framework, the Ansys R&D group is exploring the acceleration of simulation workflows with AI-based surrogate fashions. Modulus is an open-source AI framework for constructing, coaching, and fine-tuning physics-ML fashions at scale with a easy Python interface.
AI Surrogate Fashions for Actual-Time Outcomes
The NVIDIA Modulus Fourier neural operator (FNO) structure can parameterize options for a distribution of partial differential equations. Ansys researchers created an AI surrogate mannequin that effectively predicts temperature profiles for any given energy profile and an outlined flooring plan based mostly on system parameters like warmth switch coefficient, thickness, and materials properties. This mannequin presents close to real-time outcomes at considerably decreased computational prices, permitting Ansys customers to discover a wider design area for brand spanking new chips.
Following a profitable proof of idea, the Ansys group will discover the mixing of such AI surrogate fashions for its next-generation RedHawk-SC platform utilizing NVIDIA Modulus. As extra surrogate fashions are developed, the group may also look to boost mannequin generality and accuracy via in-situ fine-tuning. This may allow RedHawk-SC customers to learn from quicker simulation workflows, entry to a broader design area, and the power to refine fashions with their very own knowledge to foster innovation and security in product improvement.
To see the joint demonstration of 3D-IC multiphysics visualization utilizing NVIDIA Omniverse APIs, go to Ansys on the Design Automation Convention, working June 23-27, in San Francisco at sales space 1308 or watch the presentation on the Exhibitor Discussion board.
Picture supply: Shutterstock
