EDACafe Industry Predictions for 2024 – ANSYS

By Dr. Larry Williams, Distinguished Engineer at ANSYS Inc. 

Dr. Larry Williams

Technological Innovations

There are a number of technological innovations that I predict will turn heads in 2024.

First, continued 3D-IC and heterogeneous integration. The slowing of Dennard Scaling and Moore’s Law can be mitigated in part by 3D-IC integration, but 3D-IC has several multiphysics challenges like thermal integrity, mechanical stress, and warpage. This is due to differential thermal expansion, electromagnetic modeling for signal integrity, and full-system power integrity. These trends drive the industry to add simulation of novel physical effects as part of integrated EDA workflows. Ansys RedHawk-SC Electrothermal offers a gold-standard technology platform to examine and simulate the multiphysics behavior of 2.5D and 3D-IC designs. Designers use it to model and analyze the full-system multiphysics of entire 3D-IC designs, including chip, interposer, package, and system boundary conditions. RedHawk-SC Electrothermal relies on core solver technology from trusted golden signoff tools including RedHawk-SC™, HFSS™, Icepak™, Ansys Mechanical™, and more, to provide a comprehensive workflow for the multiphysics challenges.

From these innovations, it’s clear that 3D-IC technology is driving an inflection point in semiconductor design, which has spurred Ansys and the entire EDA industry to retool with an expanded set of multiphysics capabilities that will still be undergoing rapid advances in 2024.

Ansys is also advancing its five technology pillars—numerics, high performance computing (HPC), artificial intelligence and machine learning (AI/ML), cloud and experience, and digital engineering. Each of these pillars contributes to pushing forward technological innovations and is where we focus when driving technology development. Numerics is the core numerical algorithms we deliver to provide the most successful physics simulation. These go hand-in-hand with high performance computing, which allows simulation at an incredible scale. AI/ML is used to accelerate our simulations and to enhance user understanding of their designs. Cloud and Experience is all about building an “anywhere and anytime” ability for designers to collaborate and leverage engineering simulation. And of course, digital engineering supports a comprehensive move to a fully digital organization for product design and deployment—this includes Digital Twins.

Market Trends

I foresee a handful of market trends in 2024. A strong example is increasing product complexity. Today’s engineered products are increasingly complex, demanding new solutions for optimal design. Products have integrated electronics and semiconductors, embedded software, wired and wireless connectivity and advanced sensors and displays. However, product success requires our customers to consider the full system operation in a broad context. We’ve extended our platform to support scalable solutions that leverage new algorithms, additional physics, system solutions, embedded intelligence, HPC and integrated cloud. Our HPC product suite and cloud solutions enable enhanced insight into product performance and improve the productivity of the design process.

Bespoke silicon is another trend. Systems companies are supplying their own silicon customized for their products. They have deep systems experience that they can apply to their system on chip (SoC) solutions. We are already well on the way toward applying systems engineering concepts to designs that are vitally reliant on the quality and performance of semiconductors for their success.

Many companies are now applying a far more digital approach to product design. Model-based systems engineering (MBSE) focuses on creating and exploiting digital systems and engineering domain models as the primary means of exchange of information, feedback, and requirements, as opposed to document-centric systems engineering.

Prior to about 1990, a system engineering design was likely to be a collection of related papers and documents containing drawings, diagrams, mathematical formulas, requirements, and other specifications. Today’s systems containing electronics, electromechanics, and embedded software are too complex to manage that way, so MBSE was developed to replace static documents with “intelligent” digital models that contain everything important about the system—the requirements, the architecture, and the interfaces between the pieces of the system. Instead of paper documents that were, at best, organized into folders, these digital models are connected by a “digital thread” that can be followed to understand the entire design.

The overarching systems architect model (SAM) serves as an “authoritative source of truth” for everyone working on the project—the SAM, in which the system being designed is represented by a series of connected block diagrams to describe the system’s physical and functional architecture. The SAM must be coupled with engineering simulation to determine if the system meets requirements, which in turn, must connect to a centralized computation center. In MBSE, the combination of the SAM, CAD, and computer-aided engineering (CAE) simulation tools create the “digital thread” that links all the models and engineering data together. Ansys ModelCenter provides connection between the engineering simulation software and the SAM, enabling engineers to virtually verify and optimize their designs that may contain structural, fluids, electromagnetic, safety, and embedded software simulations.

One final market trend is connected workflow solutions. The adoption of automated flows for various applications is a significant trend we have seen within our customers—it makes sense to streamline processes, especially for repetitive workflows. This is not new to EDA, as we automate from synthesis to implementation to place-and-route, but what is new is the adoption of workflows at a higher system level. Modern workflow implementations have included automotive noise, vibration, and harshness (NVH); electromagnetic compatibility (EMC); structural fatigue; IC-specific electrothermal including IC design tools for 3D-IC; IC to package to board electrothermal; from chips to mission for aero, space, and optical; thermomechanical fatigue and durability; automated standards compliance for embedded software, functional safety, and RF exposure safety; and digital engineering transformation.

Artificial Intelligence Integration

AI is a big trending topic for 2024 and Ansys is primarily leveraging it in three ways. The first is accelerating core solvers while improving accuracy. Fluent, our flagship computational fluid dynamics (CFD) tool, uses AI to improve the accuracy of turbulence models. Ansys Mechanical and Ansys Maxwell™ predict computational spend so users can choose computing needs. Once a model is set up, the user can select “Resource Prediction” that then invokes a unique AI/ML method to predict how much memory and how long a simulation may require. This is very useful especially for selecting the correct hardware for a particular large sized simulation. The second way is enabling rapid optimization, calibration, and validation. An automotive customer in Europe is leveraging Ansys optiSLang™ machine learning to solve advanced driver assistance system challenges 1,000X faster than using Monte Carlo simulation. The third is enhancing physics-based, data-informed applications. Our hybrid digital twin solution combines AI/ML with physics simulation to predict performance of a digital twin more accurately.

In 2023, we announced three upcoming AI-infused offerings. In Q3, we announced AnsysGPT™, a multilingual, conversational virtual assistant garnering engineering expertise across physics domain. AnsysGPT will provide 24/7 comprehensive technical support for customers, delivering information and solutions more efficiently while furthering the democratization of simulation and reducing response times. In Q4, we announced that Ansys is reinforcing its ongoing investment in artificial intelligence (AI) innovation through the introduction of Ansys SimAI™ and Ansys AI+™ technologies. Ansys SimAI is a cloud-enabled, physics-neutral platform that will empower users across industries to greatly accelerate innovation and reduce time to market. With Ansys SimAI, users will be able to reliably predict performance of complex simulation scenarios rapidly. The tool will allow users to first train an AI model using simulation results and then make predictively accurate analogous designs. With Ansys AI+, Ansys will incorporate and extend AI features within its industry-leading desktop products to enhance core functionalities. The new AI+ offerings will empower customers with more choice for how they access Ansys AI capabilities across our desktop products. While Ansys SimAI will launch in early 2024, AI+ product capabilities will be available on a rolling basis beginning in Q4 2023, first with the release of Ansys optiSLang AI+™ and then Ansys Granta MI AI+™.

It’s an exciting time to be in the EDA industry and we expect a great deal of change in 2024, driven by artificial intelligence and 3D-IC technology. The adoption of AI/ML together with the maturing of 3D-IC heterogeneous integration are a perfect storm that is super-charging investment and technology development for all sectors of the semiconductor supply-chain. Product complexity, MBSE and connected workflow solutions are all necessary for upcoming trends, and simulation is key to delivering the important innovations 2024 will bring.

About Author

Dr. Larry Williams serves as Distinguished Engineer at Ansys, Inc., responsible for driving the application and direction of the company’s advanced simulation products, with emphasis on Electronics, Structural Mechanics, and Fluid Dynamics solvers. He is best known for his work on the High Frequency Structure Simulator (HFSS) for 3D high-frequency electromagnetics, antennas, and high-speed electronics. Dr. Williams is an expert in the application of electromagnetic field simulation and has over 20 years’ experience in the fields of electromagnetics and communications engineering, has delivered technical lectures internationally, and has published numerous technical papers on the subject. He serves on the UC Irvine Henry Samueli School of Engineering Dean’s Advisory Board. He received his Masters, Engineers, and Ph.D. degrees from UCLA in 1989, 1993 and 1995, respectively.

Category: Predictions

This entry was posted on Monday, January 15th, 2024 at 3:37 pm. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.