SPECIAL DIGITAL

ISSUE

AI-POWERED
ENGINEERING
 SPECIAL FOCUS
AI-POWERED ENGINEERING

Welcome
AI in Focus
Contents

4
T
GPUs Drive HPC-Powered CAE and Machine Learning he Advanced
JPR reports evolving CAE landscape. Product Develop-
ment Resource
Center (APDRC) is an
ever-growing repository of
6 GTC 2024: NVIDIA Unveils New Blackwell Processor,
Highlights Omniverse APIs and Microservices
information to help design
engineering teams take ad-
GTC ushers in new AI era with HPC chip and on-demand services. vantage of new and innova-
tive technologies that help
them create better designs
faster.
10 Dell Boosts Enterprise AI Capabilities
New Dell Technologies workstations powered by NVIDIA AI Artificial intelligence (AI) has rapidly risen
solutions unlock more engineering potential. to the top of the priority list for most enter-
prises, and engineering organizations are no
different. In simulation, AI can potentially op-
timize and accelerate analysis, while helping to

12 Autodesk AI Takes Center Stage at AU

Design and simulation workflows poised to benefit from AI
improve design productivity.
Both NVIDIA and Dell Technologies are
integration at the forefront of AI development, particu-
larly in the engineering space.
In this Special Digital Issue from APDRC,
we take a look at how artificial intelligence is

14 AIConfigurators
Drives Robotics and Automotive
at CES 2024
being incorporated into simulation and design
workflows, as well as how NVIDIA and Dell
NVIDIA delivers special address at the Consumer Electronics Show are driving this transformation.
You can learn more by visiting the resource
center at APDRC.com for the latest devel-
opments on technological advances, videos,

16 AI on Tap at 3DEXPERIENCE World 2024

Artificial intelligence to become the driving force behind
white papers and more from Digital Engineer-
ing, Dell and NVIDIA. We hope you enjoy the
SOLIDWORKS. digital issue.

Brian Albright
Editorial Director, Digital Engineering,

18 OzenCon Keynotes Highlight AI-Powered Simulation,

Digital Twins, More
Peerless Media, LLC
Comments? E-mail me at:
balbright@digitaleng.news
Conference examines the role of AI in future product development.

Editorial Staff Peerless Media, LLC

Tom Cooney Brian Ceraolo
Publisher President & CEO
PEERLESS MEDIA
Brian Albright Editorial Office
Editorial Director
Peerless Media, LLC
Paulette Chevalier 50 Speen Street, Suite 302 • Framingham, MA 01701 • 508-663-1590
Art Director de-editors@digitaleng.news • www.DigitalEngineering247.com

2 AI-Powered Engineering Summer 2024 • APDRC.com

JPR Report

GPUs Drive
In Part I, the report focuses on how GPU especially with the proliferation of servers
computing changes the workstation-based CAE outfitted with multiple high-performance
practices and applications. JPR notes, “The data center GPUs. This combination

HPC-Powered CAE
users can run larger simulations faster on their allows simulation software users to consider
desktop workstations, and as a result, they can models that were previously impractical to
optimize their designs with more iterations.” study and scenarios that were previously
In Part 1 , JPR concludes: impossible to simulate.

and Machine Learning

• GPUs can achieve 5× or more throughput Some, such as Ansys Discovery, are
for the same cost of a CPU. written from the ground up to take
• They can achieve lower cost and power advantage of the cloud and GPU. “It was
consumption for the same throughput as designed as a tool to enable engineers
the CPU. and designers to perform more upfront
JPR reports evolving CAE landscape

Image courtesy of
JPR. Image courtesy of Ansys

Part II expands the coverage to include simulations on their local workstations.

High Performance Computing (HPC) and Ansys Fellow Dipankar Choudhury
private data center and private Cloud-based described Ansys’ development of Discovery
resources, as well as Public cloud-based as a multi-physics simulation product for
resources for running large engineering design engineers,” JPR writes.

T
he Graphics Processing Unit (GPU) was once specialized hardware to simulations from vendors such as Microsoft Siemens’ Simcenter Cloud HPC and
boost visualization, prized especially by gamers. But with the introduction Azure, Rescale and ISV-specific cloud services.. STAR-CCM+ have also joined the list
of general GPU computing, championed by NVIDIA, it expanded its role. The report draws from examples in software of ISV-specific Cloud HPC solutions
Today it still makes in-game explosions and firefights much more realistic with tools from Altair, Ansys, Cadence, Dassault leveraging GPUs. Last June, Daniele Obiso,
its real-time rendering. But it has also become a way to accelerate engineering Systèmes, and Siemens. Simcenter STAR-CCM+ Technical Product
simulation, AI and machine learning. Industry analyst Jon Peddie Research’s (JPR) Manager for Siemens, wrote in a blog post,
report “Accelerating and Advancing Computer Aided Engineering Workflows” CAE at a Higher Scale “The coupled solver in Simcenter STAR-
shines a spotlight on how the GPU is changing simulation-led engineering. The A growing number of simulation software CCM+ is a very robust and efficient density-
book is authored by JPR analysts Kathleen Maher and Jon Peddie, sponsored by programs have refined their code to take based solver that has been for years the best
NVIDIA, as a free download. advantage of the GPU. Many have also added practice for several industrial applications,
tools and options to draw on on-demand HPC, amongst which: automotive vehicle external

4 AI-Powered Engineering Summer 2024 • APDRC.com APDRC.com • Summer 2024 AI-Powered Engineering 5
JPR Report

aerodynamics, aerospace aerodynamics, turbomachinery like Altair® HyperWorks®. PhysicsAI leverages historical
aero performance and Conjugate Heat Transfer (CHT) blade CAE and CAD data to predict outcomes for any physics up
cooling … In Simcenter STAR-CCM+ 2306 all this will be to 1000x faster than traditional solver simulation. This saves
available on GPU, providing you with a solution for faster organizations time and cost as engineers can test more design
turnaround time and lower costs per simulation. Moreover, we variations than ever before without the limits of parametric
ensure CPU-equivalent flow solutions by maintaining a unified studies or the need to build new simulation models.
codebase, hence providing a seamless user experience and According to benchmark data from Altair and NVIDIA,
consistent results irrespective of the hardware technology used.” using an NVIDIA RTX™ A4000 GPU provided an
8-times speedup for training physicsAI models compared
CFD Benefits from GPU to an 8-core laptop CPU
Computational Fluid Dynamics (CFD) generally puts a This January, Ansys launched Ansys SimAI, described as
heavy demand on the hardware, but it’s also shown to benefit “a physics-agnostic, software as a service (SaaS) application Image Courtesy of Koenigsegg
greatly from GPU acceleration. JPR writes, “Traditional that combines the predictive accuracy of Ansys simulation
CPU-based solvers require lengthy processing times, with the speed of generative AI.” In its announcement, the
sometimes spanning days for just a few seconds of real-world company writes, “Instead of relying on geometric parameters
activity. GPU-accelerated CFD solvers have been available to define a design, Ansys SimAI uses the shape of a design
but faced limitations in feature parity and model
size constraints. Altair’s CFD Lattice Boltzmann
Method (LBM) solver, ultraFluidX, has changed
Designed to Innovate.
Engineered for Speed.
the landscape with its efficient GPU-based
implementation, making it ideal for high-fidelity
aerodynamic and aero-acoustic simulations.”
JPR singles out the introduction of NVIDIA’s
H100 Tensor Core GPU as the watershed moment
for CFD simulation. The report says, “[It] has Enhanced with the latest advancements in AI, ray tracing,
brought about a revolution in demanding CFD and physics simulation, NVIDIA RTX™ technology enables
workloads. With up to 18,432 FP32 CUDA cores
and various configurations, it enables efficient incredible 3D designs, and photorealistic visualizations
production-scale CFD simulations with remarkable that were once unimaginable.
performance improvements.”
The NVIDIA RTX 6000 Ada GPU also has From grappling with complex simulations to crafting
18,176 CUDA cores, making workstations equipped
with the GPU viable in this space as well. immersive product experiences and more, RTX gives
you the power to tackle your most demanding
AI Gold Mine
In the last decade, NVIDIA has also been able workloads with speed and precision.
to pivot its GPUs as the preferred processors for Image courtesy Cadence
highly compute-intensive machine learning and Learn how you can accelerate your workflows and
AI algorithm development. In May 2023, The Wall Street itself as the input, facilitating broader design exploration
Journal reported, “Nvidia Joins $1 Trillion Club, Fueled by even if the structure of the shape is inconsistent across the bring your innovations to life at nvidia.com/eng-sim
AI’s Rise.” Since their introduction, NVIDIA RTX GPUs training data. The application can boost the prediction of
have included integrated Tensor Cores for accelerating AI and model performance across all design phases by 10-100X for
machine processing routines, so the same GPUs customers computation-heavy projects. Customers can train the AI using
are currently using for general compute acceleration can also previously generated Ansys or non-Ansys data. Training and
provide a significant boost for emerging AI-backed features in predictions are hosted on a state-of-the-art cloud infrastructure
engineering software. to ensure that user data is secure and kept private.”
JPR sees AI as another catalyst to drive CAE. “Employing The approach can potentially catch on, prompting other
AI and machine learning in CAE not only enables process CAE software vendors to introduce tools and features based
automation but also accelerates the development of simulation on the same principle. JPR writes, “[It] is now possible to
tools accessible to non-experts, enabling a new level of train AI on a large dataset of CAE simulations enabling
democratization in CAE. New business models are emerging evaluations beyond a component or a single design. For
to transform product development processes.” example, it’s now possible to evaluate how a variety of models
Altair has integrated their geometric deep learning engine, may interact with a variety of environments. Complexity can
Altair® physicsAI™ into user-native simulation environments increase exponentially.”

6 AI-Powered Engineering Summer 2024 • APDRC.com APDRC.com • Summer 2024 AI-Powered Engineering 7
GPU Acceleration

In March, as NVIDIA got ready to release its of our collaboration with NVIDIA, the company will

GTC 2024: NVIDIA Unveils New

Blackwell chip, Michael Dell, founder and CEO of Dell release a new product later this year—powered by NVIDIA
Technologies, said, “Dell Technologies and NVIDIA are Omniverse Cloud APIs—for Teamcenter X, our industry-
working together to shape the future of technology. With leading cloud-based product lifecycle management (PLM)

Blackwell Processor, Highlights

the launch of Blackwell, we will continue to deliver the software, part of the Siemens Xcelerator platform.”
next-generation of accelerated products and services to
our customers, providing them with the tools they need to NVIDIA NIM Inference Microservices

Omniverse APIs and Microservices

drive innovation across industries.” Dell is collaborating During GTC, NVIDIA launched NVIDIA NIM, a set
with NVIDIA to offer a next-generation compute of easy-to-use microservices for developers to create
platform based on the NVIDIA hybrid Grace ARM CPU enterprise generative AI applications. The company
with on-chip Blackwell microarchitecture B200 GPUs. describes them as “Cloud endpoints for pretrained
AI models optimized to run on hundreds of millions
Partnership with PLM, EDA of CUDA-enabled GPUs across clouds, data centers,
and CAE Vendors workstations, and PCs.” Adobe, Cadence, Getty Images,
In his keynote, Huang celebrated a number of partnerships and SAP are listed as the earliest firms to access
with leading software vendors: Cadence, Siemens , and these services, included in the NVIDIA AI Enterprise
Ansys, to name but a few. Software vendors are looking 5.0 portfolio.
to augment and infuse compute-
intensive simulation tools based
on Finite Element Analysis
(FEA) and Computational Fluid
Dynamics methods with surrogate
models and machine learning to
accelerate their solutions. The
new method speeds up simulation
but demands GPU-powered
machine learning to develop the
NVIDIA GTC, NVIDIA CEO Jensen Huang shows off Blackwell superchips combined with the necessary surrogate models.
HPC-targeted Grace CPU. Image courtesy of NVIDIA In March, Ansys announced
it will use NVIDIA Blackwell
GPUs, Hopper architecture-
based GPUs, and GB200 Grace
GTC ushers in new AI era with HPC chip and on-demand services. Blackwell Superchips to scale up
and accelerate existing solutions. During GTC, NX CAD and simulation software maker Siemens

I
n March, the 17,000-seat SAP center in San Jose, California, was filled to the brim, not by a hockey It also joined the Alliance for announces its plan to incorporate Omniverse APIs into its Xcelerator
match or a concert, but by NVIDIA GPU Technology Conference (GTC). Over the last decade, as OpenUSD (AOUSD), joining portfolio of products. Image courtesy of NVIDIA.
NVIDIA expands its focus beyond graphics processing to general purpose GPU computing and then to NVIDIA and others to promote
AI, GTC itself has also evolved to become the premiere AI conference. the use of OpenUSD 3D data interchange and framework. “These NIMs are going to help you create new types
This year, NVIDIA CEO Jensen Huang stepped up to the podium for another big reveal—the unveiling of The format is the visual language of the NVIDIA of applications for the future—not ones that you write
the new NVIDIA Blackwell chip for AI workloads. Omniverse development platform. completely from scratch, but you’re going to integrate
NVIDIA is also bolstering Omniverse’s natural language them like teams to create these applications,” Huang said.
The Blackwell Era think of themselves as one chip. Ten Terabytes of processing to remove technical barriers from potential
It’s misleading to call Blackwell a “chip,” because it’s data per second flows between them. There’s no users with NVIDIA ChatUSD and DeepSearch NIMs Humanoid Robotics
meant to function as the heart of a supercomputing memory issue, no cache issue. They function as one (NVIDIA Inference Microservices). “You can speak to During the keynote’s closing minutes, Huang was flanked
system—“a platform,” in Huang’s words. “People giant chip.” NVLink switches further streamline the [Omniverse] in English, and it would directly generate by a row of robots, from dwarfish ones to life-size ones.
think we make GPUs, and we do, but GPUs don’t data flow within the Blackwell systems. USD and talk back in USD,” Huang said. Huang said, “In the future, everything that moves will
look the way they used to,” he said. “Generative AI The Blackwell dies are connected to an NVIDIA NVIDIA announced it’s disaggregating the Omniverse be robotics. And these robotic systems, whether they are
is the defining technology of our time. Blackwell is Grace CPU, built on Arm architecture. In 2021, platform, making its core technologies available as cloud humanoid, AMRs (autonomous mobile robots), self-driving
the engine to power this new industrial revolution. NVIDIA announced its plan to build its own HPC- API. “With these APIs, you’re going to have magical cars, forklifts, or manipulating arms, they will all need one
Working with the most dynamic companies in the targeted CPU, called Grace. digital-twin capability,” Huang said. The ability to use thing—a digital twin platform. We call it Omniverse.”
world, we will realize the promise of AI for every AWS is planning to build one of the first natural language as prompts to generate 3D objects NVIDIA offers Isaac Sim, a virtual robot training
industry.” Blackwell systems, capable of processing 222 could significantly improve the user experience in CAD system. The system uses NVIDIA Omniverse’s immersive
The new Blackwell GPU comprises “208 exaFLOPS per second, Huang revealed. Others programs and quicken adoption. Therefore, CAD vendors 3D environment to replicate the robot’s operations in the
billion transistors,” Huang proudly revealed. “Two looking to build Blackwell systems include Microsoft are expected to explore these APIs. real world. The rise of robotics, Huang expects, would
dies are abutted together in such a way so they and Oracle, according to Huang. During GTC, Siemens announced, “In the next phase boost demand for Omniverse.

8 AI-Powered Engineering Summer 2024 • APDRC.com APDRC.com • Summer 2024 AI-Powered Engineering 9
AI Capabilities

Generation Laptop GPUs. According to the company, Dell AI Factory with

Dell Boosts Enterprise

All of the new devices include the Dell Optimizer for NVIDIA is an “end-to-end AI enterprise solution that
Precision AI-based software that enhances and improves integrates Dell’s compute, storage, client device, software
application and device performance based on user workflows. and services capabilities with NVIDIA’s AI infrastructure
According to a blog written by Charlie Walker, Dell and software suite, all underpinned by a high-speed

AI Capabilities
Senior Director and General Manager of the Precision and networking fabric.”
rugged product lines, the new NVIDIA RTX™ 500 and With the new Dell AI solutions, the company claims
1000 Ada Generation laptop GPUs provide the new mobile its clients will be able to more effectively and efficiently
workstations with powerful AI, ray tracing and graphics integrate AI into a wide variety of workflows.
capabilities. You can learn more about the new NVIDIA RTX The Dell AI Factory with NVIDIA supports a variety
laptop GPUs in this NVIDIA blog. of AI use cases and applications including model creation
New Dell Technologies Both the 3000 and 5000 Series mobile workstations and tuning, augmentation and inferencing. The solution
workstations “boast a combined integrated multi-processor package of can also be enhanced with professional services to help
Central Processing Unit (CPU), Neural Processing Unit customers with data preparation, implementation and
powered by NVIDIA AI (NPU) and Graphics Processing Unit (GPU), designed adoption of AI Factory.
solutions unlock to enhance the performance of AI features in more than In addition, Dell announced Dell Generative AI
more engineering
potential.

The new Dell Precision 3280 CFF and accessories. Image courtesy of Dell Technologies.

A
rtificial intelligence (AI) is rapidly being woven into engineering workflows and other types of applications.
In order to effectively utilize the technology, organizations will need to be able to corral large amounts
of data and run complex algorithms. Professional engineering workstations will play a key role in this
process. At the recent Mobile World Congress event in February and NVIDIA GTC conference in March, Dell
Technologies announced new professional-grade Precision workstations and AI-based offerings that are designed
to help engineering organizations make this transition.

“Every company that wants to remain competitive will to 8TB of RAID storage. The Dell Precision 5490 workstation. Image courtesy of Dell Technologies.
have to implement AI in some way, and AI PCs will be • The Dell Precision 3680 Tower Workstation,
central to that,” said Sam Burd, president, Client Solutions which supports up to the 14th Gen Intel Core i9K 100 applications, enabling them to operate with optimized Solutions with NVIDIA – Retrieval-Augmented
Group, Dell Technologies, in a press release. “From CPU, up to 128GB of memory, and up to 28TB of speed and power efficiency.” Generation (RAG), which leverages new
running complex AI workloads on workstations to using storage. On the GPU front, it supports up to the According to the Dell blog, this combination of processors microservices in NVIDIA AI Enterprise to offer
day-to-day AI-powered applications on laptops, the AI NVIDIA RTX 6000 Ada Generation GPU. can reduce power consumption in the laptops and improve what the company describes as a pre-validated,
­­ full-
PC will be an important investment that pays dividends in • Dell also introduced the Precision 3490 (14- performance of media-intensive AI workloads. stack solution to speed enterprise AI adoption with
productivity and paves the way to a smarter, more efficient in.), 3590 (15-in.), and 3591 (15-in.) mobile RAG, and improve GenAI model quality. Dell is also
future. Dell’s advantage starts with offering more AI PCs workstations. The 3490 and 3590 models support New AI Tools offering a full-stack solution powered by NVIDIA for
across the commercial portfolio from day one, giving the new NVIDIA RTX 500 Ada Generation A few weeks later at the NVIDIA GTC event, Dell building custom, domain-specific AI models, and a
customers the ability to start future proofing for AI today.” Laptop GPU, while the 3591 supports the announced that it is expanding its generation AI solutions to Dell Data Lakehouse for analyzing data across hybrid
At the Mobile World Congress in Barcelona, Dell NVIDIA RTX 500,RTX 1000, and RTX 2000 Ada include the new Dell AI Factory with NVIDIA. and multi cloud environments.
announced the latest versions of both its tower and mobile Generation Laptop GPUs. “Our enterprise customers are looking for an easy At GTC, Dell also announced it was working
workstations that can run complex AI workloads. • The Dell Precision 5490 (14-in.) and 5690 (16-in.) way to implement AI solutions – that is exactly what Dell with NVIDIA to introduce a rack-scale, high-density,
The new workstations include: mobile workstations support up to the Intel Ultra 9 Technologies and NVIDIA are delivering,” said Michael liquid-cooled architecture based on the NVIDIA Grace
• The Dell Precision 3280 Compact Form Factor 185H processor (16 cores), and the 5490 supports Dell, founder and CEO, Dell Technologies, as part of Blackwell Superchip, and that its Dell PowerEdge
workstation, which the company says is the world’s the NVIDIA RTX 1000, 2000 and 3000 Ada the announcement. “Through our combined efforts, XE9680 servers will support new NVIDIA Blackwell
smallest workstation that supports Tensor Core GPUs. It Generation Laptop GPUs. The Precision 5690 organizations can seamlessly integrate data with their own GPUs as well as H200 Tensor Core GPUs and NVIDIA
supports up to the Intel Core i9, NVIDIA RTX™ 4000 is available with the NVIDIA RTX 1000, RTX use cases and streamline the development of customized Quantum-2 InfiniBand and Spectrum-X Ethernet
Ada Generation GPU , 64GB of DDR5 memory, and up 2000, RTX 3500, RTX 4000, and RTX 5000 Ada GenAI models.” networking platforms.

10 AI-Powered Engineering Summer 2024 • APDRC.com APDRC.com • Summer 2024 AI-Powered Engineering 11
Autodesk AI

BlankAI, one of many technologies that comprise Autodesk AI, will debut in the automotive design studio,
enabling designers to quickly generate concepts using AI functions. Image courtesy of Autodesk Automate the Mundane, the Routine programmer, you will be given a great starting point
By now, many have heard, seen, and even used far more quickly. You’re not going to need to create
Autodesk Generative Design, part of Autodesk Fusion. everything from scratch every single time; you’re
Autodesk AI may be viewed as a more powerful, going to start at 80-90% of the way, and you’re going
more comprehensive, cross-industry approach to AI, to make tweaks, make adjustments.”
bolstered partially by Autodesk’s recent acquisition of
BlankAI, among many other things. Training Using Proprietary Data
In Day Two’s keynote, Jeff Kinder, Autodesk In an excerpt from the upcoming Autodesk State of
Executive VP of Product Development and Design & Make Spotlight on AI report, Autodesk
Manufacturing, said, “Debuting in our automotive states, “66% of industry leaders predict that AI will be
design studio next year, BlankAI will allow you to pull an integral part of their business within 2-3 years.”
from your historical library of design work, then use Many AI-based design and simulation software
it to rapidly generate new concepts that build on your makers are looking at giving users a way to further
existing design style.” refine the algorithms using proprietary data. Autodesk
According to Autodesk, BlankAI enables product is also heading in the same direction, though the exact
designers to: mechanism to facilitate it, remains unclear.
• easily generate and edit proportional 3D models In his Q&A with the industry press, Anagnost
• customize their development process with said, “At some point, as some of these capabilities get
semantic control to a level of automation, we may actually license a
• incorporate their brands’ design DNA model to a particular customer and they can train and
• dynamically navigate the data in their improve that model on top of their capabilities.
design portfolio That’s the business model that Microsoft is using
In addition, Kinder also said, “By the end of the right now for some of their tools. I think it’s a very
year, we will launch Automated Drawings in Fusion to robust model.”

Autodesk AI
translate 3D models into fully dimensioned drawings
with the click of a button.” Hardware for AI Workloads
Autodesk recently acquired FlexSim, which The use of AI/ML to train and refine algorithms
develops factory logistics and analysis software. Kinder can be data- and compute-intensive. At AU, partner

Takes Center Stage at AU

gave sneak peeks of it as part of the Autodesk Portfolio. hardware makers like Dell showcased their latest
“FlexSim powers a key link between factory design data hardware targeting such workloads.
and operational data. It helps with everything from The latest Dell Precision 7960 Tower, on display
planning and designing to building and operating a at AU, is the idea platform for AI development as
digital factory. Autodesk brings together strengths in it offers a very scalable architecture starting at 12
factory operations, building design, and construction and going up to 56 cores on the 4th Generation
solutions,” he said. Xeon W CPUs with Advanced Matrix Extensions
Design and simulation workflows poised to benefit from AI integration Autodesk also highlighted its partnership for AI acceleration. The Precision 7960 Tower with
with Cadence, a leading electronic system design the expanded chassis, as compared to the previous

T he annual Autodesk user conference, Autodesk University, was back in full swing at its usual site, The
Venetian in Las Vegas, Nevada, in November. On Day One, product designers, movie makers, game
makers, and civil engineers hurried past the flashing slot machines and roulette tables and streamed
into the Venetian Ballroom to hear Autodesk CEO Andrew Anagnost’s keynote. His message might be summed
company, to streamline smart product development
by improving collaboration across electronics and
mechanical design teams.
Kinder said, “Our integration as part of the
generation Precision 7920 Tower, can support from 1
to 4 double-wide GPUs, including the NVIDIA RTX
6000 Ada generation for maximum GPU compute
performance. Fourth-generation Tensor Cores
up by the following line. “For better or worse, AI has arrived,” said Anagnost. “We’ve been working to get you Autodesk Platform ecosystem connects Cadence’s PCB provide faster AI compute performance, delivering
excited about AI for years. But now we’re moving from talking about it to actually changing your businesses,” design capabilities directly to Fusion, enabling data to more than 2X the performance of the previous
added Anagnost. flow in both directions.” generation GPUs.
According to Scott Hamilton, Senior Industry
The change began with a change in Autodesk’s Anagnost said, “AI is the future of design and AI Speeds Up Toolpath Generation Strategist, Dell, “The extreme configurability
product lineup. As Anagnost spoke, Autodesk’s new make, and Autodesk is pioneering this transition. During the keynote, Kinder highlighted CloudNC, provides ideal options for high-end Visualization,
website for Autodesk AI went live. The company writes, We sit at the junction of many of the most creative an Autodesk partner that develops CNC automation like VRED and Omniverse use cases, and Artificial
“Autodesk AI is available to augment creative problem- and impactful industries in the world. We’ll continue software. “CloudNC add-in for AI-powered CNC Intelligence development and inferencing
solving and exploration, automate tedious and repetitive to invest in AI because of its transformational toolpath automation. CloudNC sees customers reduce applications.”
work, and give individuals and teams access to powerful potential to drive better outcomes for our customers’ tool programming times by up to 80%,” said Kinder. For more, watch the DE video report from the
analysis tools to inform decision-making.” businesses and the world.” Theo Saville, CEO of CloudNC, said, “As a CNC AU Exhibit Hall here.

12 AI-Powered Engineering Summer 2024 • APDRC.com APDRC.com • Summer 2024 AI-Powered Engineering 13
 NVIDIA’s
CES 2024
partners are
developing
automotive
configurators as warehouses and factory floors. The use of natural materials, boosting customer engagement through
powered language is already starting to transform how users create customized content,” NVIDIA points out in its
by NVIDIA and populate these dynamic environments. announcement. When running on NVIDIA Graphics
Omniverse™. Talla said, “Starting with NVIDIA PICASSO, you can Delivery Network, the automotive configurators can be
Image courtesy use a text-to-object generative AI model to generate new easily published at scale and streamed to most devices
of NVIDIA. assets and place them in the scene. With a text prompt, around the world.”
generative AI can add rain, water, rust, and more to (You can read more about one of these configurator
existing environments. Tasks that typically take an artist applications in our previous coverage here.)
days, can be completed in minutes.” NVIDIA also spotlighted in particular Mercedes-
Cobots and robots performing tasks alongside Benz, Polestar, Kodiak, and Pebble as automakers who
human operators are expected to meet regulatory have infused AI-powered features in their vehicle lines
requirements, which set the safety protocols: for and are showcasing them at the conference.
example, the speed and velocity with which they move. “Mercedes-Benz is also using digital twins for
Gerard Andrews, Senior Product Manager for Robotics, production with help from NVIDIA Omniverse,
NVIDIA, said, “We have active engagement with a a platform for developing applications to design,
number of regulatory bodies. We look at proactive collaborate, plan and operate manufacturing and

AI Drives Robotics
safety, reactive safety, and predictive safety. We are assembly facilities … Kodiak is exhibiting an
participating in efforts to drive the three concepts and autonomous truck, which relies on NVIDIA RTX
are pushing for a number of safety standards.” GPUs for high-performance compute to process the

and Automotive
Configurators at CES 2024
NVIDIA delivers special address at the Consumer Electronics Show

A
t this year’s International CES (Consumer Electronics Show, Las Vegas, Nevada, Jan. 9-12),
NVIDIA made a special address, highlighting technology that the company says will usher in a
new wave of AI-ready hardware and applications.

“Today, NVIDIA is at the center of the latest AI’s Impact on Robotics

technology transformation: generative AI,” said Jeff One area where NVIDIA expects AI-powered
Fisher, senior VP for GeForce, NVIDIA. “NVIDIA’s natural language processing to accelerate
RTX™ GPUs, with more than 100 million units developments is in the field of robotics. Deepu
shipped, are pivotal in the burgeoning field of Talla, VP of Robotics and Edge Computing,
generative AI, exemplified by innovations like NVIDIA, believes the tipping point for a robotic In robotic training, the use of natural language prompts is expected to accelerate the
ChatGPT and Stable Diffusion.” transformation is coming soon. design and creation of dynamic environments. Image courtesy of NVIDIA.
While CES caters to the consumer crowd, He said, “Generative AI models achieve
NVIDIA’s consumer-targeted initiatives are often higher accuracy while generalizing rather Omniverse-Powered Car Configurators enormous quantities of data it collects from its cameras,
precursors to what’s coming down the pike in the well. And applications are built using natural NVIDIA is also promoting its immersive 3D simulation radar and lidar sensors,” NVIDIA announced.
professional portfolios. language prompts or texts. Given these attributes, platform NVIDIA Omniverse™ to automakers at CES. In a briefing with the industry press, Danny Shapiro,
The company points out, “Mobile workstations generative AI is well-suited for robotics.” The This time, the attraction is the Omniverse-powered car VP of Automotive, NVIDIA, said, “With Omniverse-
with RTX GPUs can run NVIDIA AI Enterprise same AI model, he added, could also help generate configurators, developed and marketed by BITONE, powered configurators, consumers can experience the car
software, including NVIDIA TensorRT™ and synthetic data to augment the collected data. Brickland, Configit, Katana Studio Ltd., WPP, exactly as they want . They can customize the leather, the
NVIDIA RAPIDS™ for simplified, secure NVIDIA’s platform for training and ZeroLight, and other partners. trim, the wheels, and then take the car out for a virtual test
generative AI and data science development. A developing autonomous robots is NVIDIA Isaac “OEMs benefit from a unified asset pipeline that drive. For developers, they can even simulate the software
three-year license for NVIDIA AI Enterprise is Sim™, a robotic simulation platform. For robotic simplifies the integration of design and engineering data that runs inside the car.”
included with every NVIDIA A800 40GB Active training, in addition to the model of the robot for marketing purposes. Omniverse’s enhanced tools You can learn more in blogs that NVIDIA published
GPU, providing an ideal workstation development itself, users usually need to create a replica of also allow them to quickly produce diverse marketing during CES about its robotics and automotive capabilities.
platform for AI and data science.” the environment where the robot operates, such

14 AI-Powered Engineering Summer 2024 • APDRC.com APDRC.com • Summer 2024 AI-Powered Engineering 15
3DEXPERIENCE World 2024

and compute capabilities in a compact form factor,

offering up to 1.5x faster performance compared to the
previous-generation RTX A2000 12GB in professional

AI on Tap at workflows. And packing 16GB of GPU memory,

the RTX 2000 Ada allows professionals to handle
demanding professional workflows with ease, from

3DEXPERIENCE
using AI-powered tools to driving multi-display setups
to creating high-resolution content.
NVIDIA’s benchmark showcases that the RTX

World 2024
GPUs significantly improve performance across
The Dell Precision 3680. Image courtesy of Dell. a variety of tasks in SOLIDWORKS, including
anti-aliasing, immersive collaboration with virtual
reality (VR) models, and ray-tracing operations in
generate parametric models from picture or text input, SOLIDWORKS Visualize. NVIDIA has published the
Artificial intelligence to become and that’s what we are focusing on,” said Shrikant updated SOLIDWORKS benchmarks here: NVIDIA
the driving force behind Savant, AI and Machine Learning Expert, DS. In the RTX Accelerates SOLIDWORKS.
first release of AI-based SOLIDWORKS, “What we “The new NVIDIA RTX 2000 —with its higher-
SOLIDWORKS. want to do is give users a generative AI model that efficiency, next-generation architecture, low power
SOLIDWORKS CEO Manish Kumar outlines his vision works out of the box, but we also understand that consumption, and large frame buffer—will benefit
for how to integrate AI into the design and simulation customers would like to further customize the model SOLIDWORKS users,” said Olivier Zegdoun, Graphics
products. Image courtesy of Dassault Systemes. by training it on their own data. We’re keeping it in our Applications Research and Development Director for
mind.” SOLIDWORKS, DS. “It delivers excellent performance

I
Purchase options for SOLIDWORKS users are, as
n February, at the annual 3DEXPERIENCE World Conference (February 11-14, Dallas, Texas) hosted Kumar put it, “with the platform, or on the platform.”
by Dassault Systèmes (DS), SOLIDWORKS users came together to trade tips, network, and learn the The platform is DS’s 3DEXPERIENCE, which includes
latest features in their beloved software. The biggest takeaway was that AI would become the backbone complementary apps in the 3DEXPERIENCE Works
of SOLIDWORKS, their market-leading mechanical CAD program, and other DS products. Both Dell suite. The platform is activated once a customer begins to
Technologies and NVIDIA were also on hand with new professional-grade GPUs and workstations that can use content from the suite.
handle these AI-driven workloads. Kumar said, “Initially, there were customers who
purchased SOLIDWORKS but would not activate the
AI-Powered PCB Tools for SOLIDWORKS AI as the Backbone of 3DEXPERIENCE platform. Now, we’re seeing many of them activating
During his keynote, SOLIDWORKS CEO Manish Kumar reminded the attendees that, long before it within two days. It means they’re getting more and
Kumar invited Michael Jackson, Corporate VP of EDA ChatGPT became headline news, SOLIDWORKS more value from it.”
(Electronic Design Automation) software maker Cadence had been gradually introducing AI-driven The key difference is an “on the platform” license The NVIDIA RTX 2000 Ada Generation GPU.
Design Systems, to join him on stage. Together, the productivity tools, such as Mate Helper, Selection is a cloud-hosted browser-based SOLIDWORKS Image courtesy of NVIDIA.
two announced a partnership, bringing Cadence’s AI- Helper, and Sketch Helper. But going forward, license, allowing you to save your data to the cloud but for designers and engineers to accelerate the development
powered PCB (Printed Circuit Board) design tools to generative AI tools for model creation would take on not to your local desktop. “With the platform” means of innovative product experiences with full-model fidelity,
SOLIDWORKS users. a greater role in the design workflow. 3DEXPERIENCE apps come with a locally installed even with larger datasets.”
“SOLIDWORKS users are shifting from product- “Imagine you want to create a bookshelf. All SOLIDWORKS copy, so you can save your data to both
centric thinking to experience-centric thinking. The you have to do is provide a bounding box, and the your local drive and the cloud. A New Dell Machine
change is driving the demand for electromechanical Generative AI will create a bookshelf that fits within Bernard Charles, DS Executive Chairman of In the exhibit hall, Dell showcased the Dell Precision
devices. So today, Cadence and DS are unveiling the that box,” said Kumar. He also revealed, “To the same the Board, also made a cameo appearance during 3680 desktop tower, a recently-announced workstation
first cloud-based collaborative electromechanical AI program, we fed models of coffee tables. Now, the the keynote. Charles oversaw DS’s acquisition of targeting SOLIDWORKS users. Matthew Allard,
product, purpose-built for SOLIDWORKS,” said program could create coffee tables without manual SOLIDWORKS in 1997. At the time, SOLIDWORKS Director of Alliances and Solutions, Dell, said, “This
Kumar. intervention.” was a pioneer in Windows PC-based parametric CAD is a follow-up to our popular Dell Precision 3660.
According to the press release, the partnership will For SOLIDWORKS and other CAD software software. One of the exciting things about this machine is that
enable “collaboration for continuous development makers, the challenge is to teach the AI program “Bringing MODSIM (Modeling and Simulation) and it includes NVIDIA’s newly released RTX 2000 GPU.
across PCB, 3D mechanical design, and simulation. manufacturing constraints. Unlike consumer- AI together is an unparalleled opportunity to bring life to With an increase in video RAM (16GBs), you can now
The cloud-enabled integration offers joint customers targeted 2D text-to-image AI programs (such shapes,” he said. “Generative AI is an intrinsic part of the load larger assemblies and more complex parts.”
an easy-to-use, end-to-end solution for next-generation as Midjourney), mechanical CAD programs like fabric of the [3DEXPERIENCE] platform.” Allard also pointed out, “Increasingly, you’ll see
product development, enabling an up-to-5X reduction SOLIDWORKS need to limit the generated 3D more AI-powered features in CAD programs like
in design turnaround time.” Jackson said Cadence is objects to manufacturable topology. GPU Accelerating AI Workloads SOLIDWORKS. If these features are running locally,
launching the new tool in July. “There is no company right now that can At 3DEXPERIENCE World, NVIDIA announced they will accelerate on RTX GPUs because they have
the NVIDIA RTX™ 2000 Ada Generation GPU, Tensor processor cores targeting AI features.”
designed for AI-enhanced multi-application workflows. To learn more about configuring a workstation for
This new GPU brings cutting-edge AI, graphics SOLIDWORKS, see our guide here.

16 AI-Powered Engineering Summer 2024 • APDRC.com APDRC.com • Summer 2024 AI-Powered Engineering 17
OzenCon

software industry that is already highly consolidated workstation targeting CAD users. It can be
… The deal would bring together Synopsys’ configured with the full range of NVIDIA
semiconductor electronic design automation (EDA) professional GPUs, including NVIDIA’s newly
Ansys CTO tools with Ansys’ simulation and analysis portfolio.” released RTX 2000 Ada GPU, based on the
Prith Banerjee A few weeks before OzenCon, at the NVIDIA Ada Lovelace GPU architecture, featuring
discusses the 3DEXPERIENCE World event in Las Vegas, 16GB of high performance GPU memory.
integration Synopsys’s rival Cadence announced a partnership Manufacturing workflows typically include
of hardware, with popular CAD software maker SOLIDWORKS design and simulation for virtual prototyping that
software, and to bring their solutions together. These acquisitions require desktop workstations with GPUs for quick
electronics in and partnerships foretell future corporate handshakes and realistic visualization of intricate geometries
future products. that will bring chip design and mechanical design and for rapid computations.
Photo by Kenneth closer than ever. Zihan Wang, High-Tech Strategy and
Wong, DE 24/7. Ansys is also injecting AI into its simulation tools, Operations manager for NVIDIA, said, “If you
aiming to make them run faster and become easier to want to simulate the motion of granular materials
use. “We just announced a tool earlier this year called in the mining industry, like particles on a conveyor
Ansys SimAI, which lets you take any of our black-box belt in a mixer, double-precision GPUs are required
solvers and train an AI model. Once trained, it will run to speed up the simulation. We have the NVIDIA
simulation hundreds of times faster,” said Banerjee. RTX A800 40GB Active GPU for that type of

OzenCon Keynotes Highlight

Different Twins for Different Needs simulation. It’s a workstation GPU so you can
Teresa Alberts, CEO of ITficient, was another embed it into your Dell Precision workstation …
keynote speaker at OzenCon. Her talk was titled, For simulations where single precision provides the

AI-Powered Simulation,
“Applications and Benefits of Digital Twin for the required accuracy, NVIDIA RTX Ada Generation
Manufacturing Industry.” Alberts said, “Where the GPUs will provide incredible performance.”
digital twin brings value depends on the product As AI speeds up processing, Wang expects
and what the customer’s pain points are. If they can engineers to expand the use of simulation. “With

Digital Twins, More identify the pain points, then the digital twin can help
address them, and bring added value.”
Alberts highlighted the case of several power
plant operators in her presentation. “They use
AI-embedded HPC, you can process more data
points to make a better decision,” he said.

digital twins to ensure the plants’ availability, to

Conference examines the role of AI in future see the condition of the critical components, to
product development. get forecasts about the remaining lifetimes of
the critical components, and to service them in
advance.”

O
zenCon, the simulation user conference hosted by Ansys reseller Ozen Engineering Inc., For digital twin owners and operators, the
took place at the Computer History Museum (Mountain View, California) on February amount of data collected from sensors would be a
22. Dr. Prith Banerjee, CTO of Ansys, joined the conference as the keynote speaker. significant data analysis burden. Ansys anticipates
His talk entitled “The Future of Digital Engineering,” acknowledged that while the use of its tool can fill a need. “By integrating AI-based
simulation and test to develop products had been the norm for the past 50 years or so, a sea data analytics with simulation, you can create what
change was rapidly approaching, according to him. we call hybrid digital twins.”

“Digital engineering is how you engineer and so on,” said Banerjee. AI-Ready Hardware
future products, which include hardware, Many of the AI workloads, such as using machine
mechanical and electronics parts, as well as Merging ECAD and MCAD learning to perform data analysis or object
software. Typically, these things are done The silos between electrical and recognition, are expected to drive up the demand
in a siloed way, with different teams doing mechanical are evaporating, due in large for computing power, often beyond what’s
different things … Ansys is preparing part to partnerships between ECAD and normally available in a personal workstation. Most
engineers to use requirements, written MCAD vendors, as well as from industry likely, engineering simulation in the future will
in languages like SysML, to use a system consolidation. In January, Synopsys signed a require a mix of on-premise hardware and on-
architecture modeling tool called SAM, and to deal to acquire Ansys for $35 billion. Reuters demand computing. Teresa Alberts, CEO of ITficient, discusses the reason
connect that with our accurate 3D solvers like reported, “The transaction will create a This month, Dell announced the new different people use digital twins. Photo by Kenneth
Ansys Mechanical, LS-DYNA, Ansys Fluent, massive new player in a sector of the business Dell Precision 3680 desktop tower, the latest Wong, DE 24/7.

18 AI-Powered Engineering Summer 2024 • APDRC.com APDRC.com • Summer 2024 AI-Powered Engineering 19