Monthly Archives: April 2021

Applying for a home mortgage can resemble a part-time job. But whether consumers are seeking out a home loan, car loan or credit card, there’s an incredible amount of work going on behind the scenes in a bank’s decision — especially if it has to say no.

To comply with an alphabet soup of financial regulations, banks and mortgage lenders have to keep pace with explaining the reasons for rejections to both applicants and regulators.

Busy in this domain, Wells Fargo will present at NVIDIA GTC21 this week some of its latest development work behind this complex decision-making using AI models accelerated by GPUs.

To inform their decisions, lenders have historically applied linear and non-linear regression models for financial forecasting and logistic and survivability models for default risk. These simple, decades-old methods are easy to explain to customers.

But machine learning and deep learning models are reinventing risk forecasting and in the process requiring explainable AI, or XAI, to allow for customer and regulatory disclosures.

Machine learning and deep learning techniques are more accurate but also more complex, which means banks need to spend extra effort to be able to explain decisions to customers and regulators.

These more powerful models allow banks to do a better job understanding the riskiness of loans, and may allow them to say yes to applicants that would have been rejected by a simpler model.

At the same time, these powerful models require more processing, so financial services firms like Wells Fargo are moving to GPU-accelerated models to improve processing, accuracy and explainability, and to provide faster results to consumers and regulators.

What Is Explainable AI?

Explainable AI is a set of tools and techniques that help understand the math inside an AI model.

XAI maps out the data inputs with the data outputs of models in a way that people can understand.

“You have all the linear sub-models, and you can see which factor is the most significant — you can see it very clearly,” said Agus Sudjianto, executive vice president and head of Corporate Model Risk at Wells Fargo, explaining his team’s recent work on Linear Iterative Feature Embedding (LIFE) in a research paper.

Wells Fargo XAI Development

The LIFE algorithm was developed to handle high prediction accuracy, ease of interpretation and efficient computation.

LIFE outperforms directly trained single-layer networks, according to Wells Fargo, as well as many other benchmark models in experiments.

The research paper — titled Linear Iterative Feature Embedding: An Ensemble Framework for Interpretable Model — authors include Sudjianto, Jinwen Qiu, Miaoqi Li and Jie Chen.

Default or No Default 

Using LIFE, the bank can generate codes that correlate to model interpretability, offering the right explanations to which variables weighed heaviest in the decision. For example, codes might be generated for high debt-to-income ratio or a FICO score that fell below a set minimum for a particular loan product.

There can be anywhere from 40 to 80 different variables taken into consideration for explaining rejections.

“We assess whether the customer is able to repay the loan. And then if we decline the loan, we can give a reason from a recent code as to why it was declined,” said Sudjianto.

Future Work at Wells Fargo

Wells Fargo is also working on Deep ReLU networks to further its efforts in model explainability. Two of the team’s developers will be discussing research from their paper, Unwrapping The Black Box of Deep ReLU Networks: Interpretability, Diagnostics, and Simplification, at GTC.

Learn more about the LIFE model work by attending the GTC talk by Jie Chen, managing director for Corporate Model Risk at Wells Fargo. Learn about model work on Deep ReLU Networks by attending the talk by Aijun Zhang, a quantitative analytics specialist at Wells Fargo, and Zebin Yang, a Ph.D. student at Hong Kong University. 

Registration for GTC is free.

Image courtesy of joão vincient lewis on Unsplash

NVIDIA technology has been behind some of the world’s most stunning virtual reality experiences.

Each new generation of GPUs has raised the bar for VR environments, producing interactive experiences with photorealistic details to bring new levels of productivity, collaboration and fun.

And with each GTC, we’ve introduced new technologies and software development kits that help developers create extended reality (XR) content and experiences that are more immersive and delightful than ever.

From tetherless streaming with NVIDIA CloudXR to collaborating in a virtual world with NVIDIA Omniverse, our latest technologies are powering the next generation of XR.

This year at GTC, NVIDIA announced a new release for CloudXR that adds support for iOS. We also had announcements with leading cloud service providers to deliver high-quality XR streaming from the cloud. And we released a new version of Variable Rate Supersampling to improve visual performance.

Bringing High Performance and VR Mobility Together

NVIDIA CloudXR is an advanced technology that gives XR users the best of both worlds: the performance of NVIDIA GPUs with the mobility of untethered all-in-one head-mounted displays.

CloudXR is designed to stream all kinds of XR content from any server to any device. Users can easily access powerful, high-quality immersive experiences from anywhere in the world, without being physically connected to a workstation.

From product designers reviewing 3D models to first responders running training simulations, anyone can benefit from CloudXR using Windows and Android devices. We will soon be releasing CloudXR 2.1, which adds support for Apple iOS AR devices, including iPads and iPhones.

Taking XR Streaming to the Cloud

With 5G networks rolling out, streaming XR over 5G from the cloud has the potential to significantly enhance workflows across industries. But the big challenge with delivering XR from the cloud is latency — for people to have a great VR experience, they have to maintain 20ms motion-to-photon latency.

To deliver the best cloud streaming experience, we’ve fine-tuned NVIDIA CloudXR. Over the past six months, we’ve taken great strides to bring CloudXR streaming to cloud service providers, from Amazon Web Services to Tencent.

This year at GTC, we’re continuing this march forward with additional news:

• CloudXR is available on Amazon Marketplace as an Amazon Machine Image (AMI)
• The CloudXR platform will be available on NVIDIA GPU-powered virtual machine instances on Azure

Also at GTC, Google will present a session that showcases CloudXR running on a Google Cloud instance.

To support CloudXR everywhere, we’re adding more client devices to our family.

We’ve worked with Qualcomm Technologies to deliver boundless XR, and with Ericsson on its 5G radio and packet core infrastructure to optimize CloudXR. Hear about the translation of this work to the manufacturing environment at BT’s session in GTC’s XR track.

And we’ve collaborated with Magic Leap on a CloudXR integration, which they will present at GTC. Magic Leap and CloudXR provide a great step forward for spatial computing and an advanced solution that brings many benefits to enterprise customers.

Redefining the XR Experience 

The quality of visuals in a VR experience is critical to provide users with the best visual performance. That’s why NVIDIA developed Variable Rate Supersampling (VRSS), which allows rendering resources to be focused in a foveated region where they’ll have the greatest impact on image quality.

The first VRSS version supported fixed foveated rendering in the center of the screen. The latest version, VRSS 2, integrates dynamic gaze tracking, moving the foveated region where the user is looking.

These advances in XR technology are also paving the way for a solution that allows users to learn, work, collaborate or play with others in a highly realistic immersive environment. The CloudXR iOS integration will soon be available in NVIDIA Omniverse, a collaboration and simulation platform that streamlines 3D production pipelines.

Teams around the world can enter Omniverse and simultaneously collaborate across leading content creation applications in a shared virtual space. With the upcoming CloudXR 2.1 release, Omniverse users can stream specific AR solutions using their iOS tablets and phones.

Expanding XR Workflows at GTC

Learn more about these advances in XR technology at GTC. Register for free and explore over 40 speaker sessions that cover a variety of XR topics, from NVIDIA Omniverse to AI integrations.

Check out the latest XR demos, and get access to an exclusive Connect with Experts session.

And watch a replay of the GTC keynote address by NVIDIA CEO Jensen Huang to catch up on the latest announcements.

Sign up to get news and updates on NVIDIA XR technologies.

Feature image credit: Autodesk VRED.

Breakthroughs in 3D model visualization, such as real-time ray–traced rendering and immersive virtual reality, are making architecture and design workflows faster, better and safer.  

At GTC this week, NVIDIA announced the newest advances for the AEC industry with the latest NVIDIA Ampere architecture-based enterprise desktop RTX GPUs, along with an expanded range of mobile laptop GPUs.  

AEC professionals will also want to learn more about NVIDIA Omniverse Enterprise, an open platform for 3D collaboration and physically accurate simulation. 

New RTX GPUs Bring More Power, Performance for AEC 

The NVIDIA RTX A5000 and A4000 GPUs are designed to enhance workflows for architectural design visualization. 

Based on NVIDIA Ampere architecture, the RTX A5000 and A4000 integrate second–generation RT Cores to further boost ray tracing , and third–generation Tensor Cores to accelerate AI-powered workflows such as rendering denoising, deep learning super sampling and generative design.  

Several architecture firms, including HNTB, have experienced how the RTX A5000 enhances design workflows.  

“The performance we get from the NVIDIA RTX A5000, even when enabling NVIDIA RTX Global Illumination, is amazing,” said Austin Reed, director of creative media studio​ at HNTB. “Having NVIDIA RTX professional GPUs at our designers’ desks at HNTB will enable us to fully leverage RTX technology in our everyday workflows.”  

NVIDIA’s new range of mobile laptop GPU models — including the NVIDIA RTX A5000, A4000, A3000 and A2000, and the NVIDIA T1200, T600 and T500 — allows AEC professionals to select the perfect GPU for their workloads and budgets.  

With this array of choices, millions of AEC professionals can do their best work from anywhere, even compute-intensive work such as immersive VR for construction rehearsals or point cloud visualization of massive 3D models.  

NVIDIA Omniverse Enterprise: A Shared Space for 3D Collaboration  

Architecture firms can now accelerate graphics and simulation workflows with NVIDIA Omniverse Enterprise, the world’s first technology platform that enables global 3D design teams to simultaneously collaborate in a shared virtual space. 

The platform enables organizations to unite their assets and design software tools, so AEC professionals can collaborate on a single project file in real time. 

Powered by NVIDIA RTX technology, Omniverse delivers high-performance and physically accurate simulation for complex 3D scenes like cityscapes, along with real-time ray– and path–traced rendering. Architects and designers can instantly share physically accurate models across teams and devices, accelerating design workflows and reducing the number of review cycles.  

Artists Create Futuristic Renderings with NVIDIA RTX  

Overlapping with GTC, the “Building Utopia” design challenge allowed archviz specialists around the world to discover how NVIDIA RTX real-time rendering is transforming architectural design visualization. 

Our thanks to all the participants who showcased their creativity and submitted short animations they generated using Chaos Vantage running on NVIDIA RTX GPUs. NVIDIA, Lenovo, Chaos Group, KitBash3D and CG Architect are thrilled to announce the winners. 

Congratulations to the winner, Yi Xiang, who receives a Lenovo ThinkPad P15 with an NVIDIA Quadro RTX 5000 GPU. In second place, Cheng Lei will get an NVIDIA Quadro RTX 8000, and in third place, Dariele Polinar will receive an NVIDIA Quadro RTX 6000. 

Discover More AEC Content at GTC 

Learn more about the newest innovations and all the AEC-focused content at GTC by registering for free.  

Check out the latest GTC demos that showcase amazing technology. Join sessions on NVIDIA Omniverse presented by leading architecture firms like CannonDesign, KPF and Woods Bagot. Learn how companies like The Grid Factory and The Gettys Group are using RTX-powered immersive experiences to accelerate design workflows.  

And be sure to watch a replay of the GTC keynote address by NVIDIA founder and CEO Jensen Huang.

Featured image courtesy of KPF – Beijing Century City – 北京世纪城市.

The factories of the future will have a soul — a “digital twin” that blends man and machine in stunning new ways.

In a demo blending reality and virtual reality, robotics and AI, to manage one of BMW’s automotive factories, NVIDIA CEO Jensen Huang Monday rolled out a stunning vision of the future of manufacturing.

“We are working with BMW to create a future factory,” Huang announced during his keynote address at NVIDIA’s GPU Technology Conference before giving his audience a look.

The demo highlights the general availability of NVIDIA Omniverse Enterprise, the first technology platform enabling global 3D design teams to work together simultaneously across multiple software suites in a shared virtual space.

The AI factory demo brings a full suite of NVIDIA technologies on Omniverse, including the NVIDIA Isaac platform for robotics, the NVIDIA EGX edge computing platform and the NVIDIA Aerial software development kit, which brings GPU-accelerated, software-defined 5G wireless radio access networks to the factory floor.

‘The World’s Largest Custom-Manufacturing Company’

Each of BMW’s factory lines can produce up to 10 different cars, and BMW prides itself on giving customers plenty of choices.

There are over 100 options for each car, and more than 40 BMW models. In all, there are 2,100 possible ways to configure a new BMW.

“BMW may very well be the world’s largest custom-manufacturing company,” Huang said.

These vehicles are produced in 31 factories located around the world, explained Milan Nedeljković, member of the Board of Management of BMW AG.

Moving the Parts That Go into the Machines That Move Your Parts

In an instant, Huang and Nedeljković summoned a digital twin of one of BMW’s factories — and the screen was filled with gleaming cars being assembled by banks of perfectly synchronized robots — all simulated.

To design and reconfigure its factories, BMW’s global teams can collaborate in real-time using different software packages like Revit, Catia, or point clouds to design and plan the factory in 3D and all the changes are visible, in real-time, on Omniverse.

“The capability to operate in a perfect simulation revolutionalizes BMW’s planning processes,” Nedeljković said.

Some of that work has to be hands-on. BMW regularly reconfigures our factories to accommodate new vehicle launches. Now, thanks to Omniverse that doesn’t mean workers have to travel.

Nedeljković showed two BMW planning experts located in different parts of the world testing a new line design in Omniverse.

One of them “wormholes” — or travels virtually — into an assembly simulation with a motion capture suit and records task movements.

The other adjusts the line design, in real time.

“They work together to optimize the line as well as worker ergonomics and safety,” Nedeljković said.

The next step: recreating these kinds of interactions, at scale, in simulations, Nedeljković said.

“That’s exactly why NVIDIA has Digital Human for simulation,” Huang said. “Digital Humans are trained with data from real associates.”

These digital humans can be used in simulations to test new workflows for worker ergonomics and efficiency.

BMW’s 57,000 factory workers share workspace with robots designed to make their jobs easier.

Omniverse, Nedeljković said, will help robots adapt to BMW’s reconfigured factories rapidly.

“With NVIDIA Isaac robotics platform, BMW is deploying a fleet of intelligent robots for logistics to improve the material flow in our production,” Nedeljković said.

That agility is necessary since BMW produces 2.5 million vehicles per year, and 99 percent of them are custom.

Omniverse can tap into NVIDIA Isaac for synthetic data generation and domain randomization, Huang said. That’s key to bootstrapping machine learning.

“Isaac Sim generates millions of relevant synthetic images, and varies the environment to teach robots. ” Huang said.

Domain randomization can generate an infinite permutation of photorealistic objects, textures, orientations, and lighting conditions, Huang said.

“Simulation offers perfect ground truth, whether for detection, segmentation or depth perception,” he added.

Huang and Nedeljković showed a BMW employee monitoring operations in the factory. The operator is able to assign missions to different robots, and see a photorealistic digital win of its progress in Omniverse — all updated by sensors throughout the factory.

With NVIDIA Fleet Command software, workers can securely orchestrate robots, and other devices, in the factory, Huang explained.

They can monitor complex manufacturing cells in real-time, update software over the air, and launch robots in the factory on missions.

Humans can even lend robots a “helping hand.” When an alert is sent to MIssion Control, one of BMW”s human associations can teleoperate the robot — looking through its camera to guide it through a 5G connection.

Then, with a push of a button, the operator returns the robot to autonomous control.

Continuous Improvement, Continually Improving

“NVIDIA Omniverse and NVIDIA AI give us the chance to simulate the 31 factories in our production network,” Nedeljković said.

All the elements of the complete factory model — including the associates, the robots, the buildings, the assembly parts — can be simulated to support a wide range of AI-enabled use cases such as virtual factory planning, autonomous robots, predictive maintenance, big data analytics, he explained.

“These new innovations will reduce the planning times, improve flexibility and precision, and at the end produce 30 percent more efficient planning processes,” Nedeljković said.

The result: a beautifully crafted new car, an amazing machine that’s the product of an amazing machine — a factory able to capture and replicate every motion in the real world to a digital one, and back.

The electric vehicle revolution is about to reach the next level.

Leading startups and EV brands have all announced plans to deliver intelligent vehicles to the mass market beginning in 2022. And these new, clean-energy fleets will achieve AI capabilities for greater safety and efficiency with the high-performance compute of NVIDIA DRIVE.

The car industry has become a technology industry — future cars will be completely programmable with software-driven business models. Companies will offer services and subscriptions over the air for the life of the car.

These new energy vehicles, or NEVs, will kickstart this transition with centralized, software-defined compute that enables continuously improving, cutting-edge AI capabilities.

NEV Newcomers

For some companies, 2022 marks the initial launch of their visionary concepts into production reality.

Canoo unveiled its first vehicle — an all-electric van — in 2019. Now, the startup is on track to deliver an entire line of EVs, including a delivery truck, pickup and sports sedan, to customers starting in 2022.

Canoo’s flagship personal vehicle will leverage NVIDIA DRIVE AGX Xavier for smart driver assistance features. And since the DRIVE AGX platform is open and scalable, Canoo can continue to develop increasingly advanced capabilities through the life of its vehicles.

Also on the horizon is the much anticipated Faraday Future FF91. This premium EV is designed to be an intelligent third living space, with a luxurious interior packed with convenience features powered by NVIDIA DRIVE.

Also charging onto the EV scene is Vinfast, a startup planning to launch a fleet of smart vehicles beginning in 2022. These vehicles will provide industry-leading safety and enhanced autonomy, leveraging the AI compute of NVIDIA DRIVE Xavier, and for proceeding generations, NVIDIA DRIVE Orin.

“NVIDIA is a vital partner for our work in autonomous driving,” said Hung Bui, chief executive of VinAI. “NVIDIA DRIVE delivers the core compute for our vehicles, delivering advanced sensing and other expanding capabilities.”

A Leading Legacy

NIO has announced a supercomputer to power its automated and autonomous driving features, with NVIDIA DRIVE Orin at its core.

The computer, known as Adam, will achieve over 1,000 trillion operations per second of performance with the redundancy and diversity necessary for safe autonomous driving. It also enables personalization in the vehicle, learning from individual driving habits and preferences while continuously improving from fleet data.

The Orin-powered supercomputer will debut in the flagship ET7 sedan, scheduled for production in 2022, and will be in every NIO model to follow.

Breakout EV maker Li Auto will also develop its next generation of electric vehicles using NVIDIA DRIVE AGX Orin. These new vehicles are being developed in collaboration with tier 1 supplier Desay SV and feature advanced autonomous driving features, as well as extended battery range for truly intelligent mobility.

This high-performance platform will enable Li Auto to deploy an independent, advanced autonomous driving system with its upcoming fleet.

Xpeng is already putting its advanced driving technology on the road. In March, the automaker completed a six-day cross-country autonomous drive with a fleet of intelligent P7 sedans. The vehicles operated without human intervention using the XPilot 3.0 autonomous driving system, powered by NVIDIA DRIVE AGX Xavier.

Finally, one of the world’s largest automakers, SAIC, is evolving to meet the industry’s biggest technological transformations with two new EV brands packed with advanced AI features.

R-Auto is a family of next-generation vehicles featuring the R-Tech advanced intelligent assistant, powered by NVIDIA DRIVE AGX Orin. R-Tech uses the unprecedented level of compute performance of Orin to run perception, sensor fusion and prediction for automated driving features in real time.

The ultra-premium IM brand is the product of a partnership with etail giant Alibaba. The long-range electric vehicles will feature AI capabilities powered by the high-performance, energy-efficient NVIDIA DRIVE Orin compute platform.

The first two vehicles in the lineup — a flagship sedan and SUV — will have autonomous parking and other automated driving features, as well as a 93kWh battery that comes standard. SAIC will begin taking orders for the sedan at the Shanghai Auto Show later this month, with the SUV following in 2022.

EVs are driving the next decade of transportation. And with NVIDIA DRIVE at the core, these vehicles have the intelligence and performance to go the distance.

Buckle up. NVIDIA CEO Jensen Huang just laid out a singular vision filled with autonomous machines, super-intelligent AIs and sprawling virtual worlds – from silicon to supercomputers to AI software – in a single presentation.

“NVIDIA is a computing platform company, helping to advance the work for the Da Vincis of our time – in language understanding, drug discovery, or quantum computing,” Huang said in a talk delivered from behind his kitchen counter to NVIDIA’s GPU Technology Conference. “NVIDIA is the instrument for your life’s work.”

During a presentation punctuated with product announcements, partnerships, and demos that danced up and down the modern technology stack, Huang spoke about how NVIDIA is investing heavily in CPUs, DPUs, and GPUs and weaving them into new data center scale computing solutions for researchers and enterprises.

He talked about NVIDIA as a software company, offering a host of software built on NVIDIA AI as well as NVIDIA Omniverse for simulation, collaboration, and training autonomous machines.

Finally, Huang spoke about how NVIDIA is moving automotive computing forward with a new SoC, NVIDIA Atlan, and new simulation capabilities.

CPUs, DPUs and GPUs

Huang announced NVIDIA’s first data center CPU, Grace, named after Grace Hopper, a U.S. Navy rear admiral and computer programming pioneer.

Grace is a highly specialized processor targeting largest data intensive HPC and AI applications as the training of next-generation natural-language processing models that have more than one trillion parameters.

When tightly coupled with NVIDIA GPUs, a Grace-based system will deliver 10x faster performance than today’s state-of-the-art NVIDIA DGX-based systems, which run on x86 CPUs.

While the vast majority of data centers are expected to be served by existing CPUs, Gracewill serve a niche segment of computing.“Grace highlights the beauty of Arm,” Huang said.

Huang also announced that the Swiss National Supercomputing Center will build a supercomputer, dubbed Alps, will be powered by Grace and NVIDIA’s next-generation GPU. U.S. Department of Energy’s Los Alamos National Laboratory will also bring a Grace-powered supercomputer online in 2023, NVIDIA announced.

Accelerating Data Centers with BlueField-3

Further accelerating the infrastructure upon which hyperscale data centers, workstations, and supercomputers are built, Huang announced the NVIDIA BlueField-3 DPU.

The next-generation data processing unit will deliver the most powerful software-defined networking, storage and cybersecurity acceleration capabilities.

Where BlueField-2 offloaded the equivalent of 30 CPU cores, it would take 300 CPU cores to secure, offload, and accelerate network traffic at 400 Gbps as BlueField-3— a 10x leap in performance, Huang explained.

‘Three Chips’

Grace and BlueField are key parts of a data center roadmap consisting of 3 chips: CPU, GPU, and DPU, Huang said. Each chip architecture has a two-year rhythm with likely a kicker in between. One year will focus on x86 platforms, the next on Arm platforms.

“Every year will see new exciting products from us,” Huang said. “Three chips, yearly leaps, one architecture.”

Expanding Arm into the Cloud 

Arm, Huang said, is the most popular CPU in the world. “For good reason – it’s super energy-efficient and its open licensing model inspires a world of innovators,” he said.

For other markets like cloud, enterprise and edge data centers, supercomputing, and PC, Arm is just starting. Huang announced key Arm partnerships — Amazon Web Services in cloud computing, Ampere Computing in scientific and cloud computing, Marvel in hyper-converged edge servers, and MediaTek to create a Chrome OS and Linux PC SDK and reference system.

DGX – A Computer for AI

Weaving together NVIDIA silicon and software, Huang announced upgrades to NVIDIA’s DGX Station “AI data center in-a-box” for workgroups, and the NVIDIA DGX SuperPod, NVIDIA’s AI-data-center-as-a-product for intensive AI research and development.

The new DGX Station 320G harnesses 320Gbytes of super-fast HBM2e connected to 4 NVIDIA A100 GPUs over 8 terabytes per second of memory bandwidth. Yet it plugs into a normal wall outlet and consumes just 1500 watts of power, Huang said.

The DGX SuperPOD gets the new 80GB NVIDIA A100, bringing the SuperPOD to 90 terabytes of HBM2e memory. It’s been upgraded with NVIDIA BlueField-2, and NVIDIA is now offering it with the NVIDIA Base Command DGX management and orchestration tool.

NVIDIA EGX for Enterprise 

Further democratizing AI, Huang introduced a new class of NVIDIA-certified systems, high-volume enterprise servers from top manufacturers. They’re now certified to run the NVIDIA AI Enterprise software suite, exclusively certified for VMware vSphere 7, the world’s most widely used compute virtualization platform.

Expanding the NVIDIA-certified servers ecosystem is a new wave of systems featuring the NVIDIA A30 GPU for mainstream AI and data analytics and the NVIDIA A10 GPU for AI-enabled graphics, virtual workstations and mixed compute and graphics workloads, announced today.

AI-on-5G

Huang also discussed NVIDIA’s AI-on-5G computing platform – bringing together 5G and AI into a new type of computing platform designed for the edge that pairs the NVIDIA Aerial software development kit with the NVIDIA BlueField-2 A100, combining GPUs and CPUs into “the most advanced PCIE card ever created.”

Partners Fujitsu, Google Cloud, Mavenir, Radisys and Wind River are all developing solutions for NVIDIA’s AI-on-5G platform.

NVIDIA AI and NVIDIA Omniverse

Virtual, real-time, 3d worlds inhabited by people, AIs, and robots are no longer science-fiction.

NVIDIA Omniverse is cloud-native, scalable to multiple GPUs, physically accurate, takes advantage of RTX real-time path tracing and DLSS, simulates materials with NVIDIA MDL, simulates physics with NVIDIA PhysX, and fully integrates NVIDIA AI, Huang explained.

“Omniverse was made to create shared virtual 3D worlds,” Huang said. “Ones not unlike the science fiction metaverse described by Neal Stephenson in his early 1990s novel ‘Snow Crash’”

Huang announced that starting this summer, Omniverse will be available for enterprise licensing. Since its release in open beta partners such as Foster and Partners in architecture, ILM in entertainment, Activision in gaming, and advertising powerhouse WPP have put Omniverse to work.

The Factory of the Future

To show what’s possible with Omniverse Huang, along with Milan Nedeljković, member of the Board of Management of BMW AG, showed how a photorealistic, real-time digital model — a “digital twin” of one of BMW’s highly-automated factories — can accelerate modern manufacturing.

“These new innovations will reduce the planning times, improve flexibility and precision and at the end produce 30 percent more efficient planning,” Nedeljković said.

A Host of AI Software

Huang announced NVIDIA Megatron — a framework for training Transformers, which have led to breakthroughs in natural-language processing. Transformers generate document summaries, complete phrases in email, grade quizzes, generate live sports commentary, even code.

He detailed new models for Clara Discovery — NVIDIA’s acceleration libraries for computational drug discovery, and a partnership with Schrodinger — the leading physics-based and machine learning computational platform for drug discovery and material science.

To accelerate research into quantum computing — which relies on quantum bits, or qubits, that can be 0, 1, or both — Huang introduced cuQuantum to accelerate quantum circuit simulators so researchers can design better quantum computers.

To secure modern data centers, Huang announced NVIDIA Morpheus – a data center security platform for real-time all-packet inspection built on NVIDIA AI, NVIDIA BlueField, Net-Q network telemetry software, and EGX.

To accelerate conversational AI, Huang announced the availability of NVIDIA Jarvis – a state-of-the-art deep learning AI for speech recognition, language understanding, translations, and expressive speech.

To accelerate recommender systems — the engine for search, ads, online shopping, music, books, movies, user-generated content, and news — Huang announced NVIDIA Merlin is now available on NGC, NVIDIA’s catalog of deep learning framework containers.

And to help customers turn their expertise into AI, Huang introduced NVIDIA TAO to fine-tune and adapt NVIDIA pre-trained models with data from customers and partners while protecting data privacy.

“There is infinite diversity of application domains, environments, and specializations,” Huang said. “No one has all the data – sometimes it’s rare, sometimes it’s a trade secret.

The final piece is the inference server, NVIDIA Triton, to glean insights from the continuous streams of data coming into customer’s EGX servers or cloud instances, Huang said.

‘Any AI model that runs on cuDNN, so basically every AI model,” Huang said. “From any framework – TensorFlow, Pytorch, ONNX, OpenVINO, TensorRT, or custom C++/python backends.”

Advancing Automotive with NVIDIA DRIVE

Autonomous vehicles are “one of the most intense machine learning and robotics challenges – one of the hardest but also with the greatest impact,” Huang said.

NVIDIA is building modular, end-to-end solutions for the $10 trillion transportation industry so partners can leverage the parts they need.

Huang said NVIDIA DRIVE Orin, NVIDIA’s AV computing system-on-a-chip, which goes into production in 2022, was designed to be the car’s central computer.

Volvo Cars has been using the high-performance, energy-efficient compute of NVIDIA DRIVE since 2016 and developing AI-assisted driving features for new models on NVIDIA DRIVE Xavier with software developed in-house and by Zenseact, Volvo Cars’ autonomous driving software development company.

And Volvo Cars announced during the GTC keynote today that it will use NVIDIA DRIVE Orin to power the autonomous driving computer in its next-generation cars.

The decision deepens the companies’ collaboration to even more software-defined model lineups, beginning with the next-generation XC90, set to debut next year.

Meanwhile, NVIDIA DRIVE Atlan, NVIDIA’s next-generation automotive system-on-a-chip, and a true data center on wheels, “will be yet another giant leap,” Huang announced.

Atlan will deliver more than 1,000 trillion operations per second, or TOPS, and targets 2025 models.

“Atlan will be a technical marvel – fusing all of NVIDIA’s technologies in AI, auto, robotics, safety, and BlueField secure data centers,” Huang said.

Huang also announced the NVIDIA 8th generation Hyperion car platform – including reference sensors, AV and central computers, 3D ground-truth data recorders, networking, and all of the essential software.

Huang also announced that DRIVE Sim will be available for the community this summer.

Just as Omniverse can build a digital twin of the factories that produce cars, DRIVE Sim can be used to create a digital twin of autonomous vehicles to be used throughout AV development.

“The DRIVE digital twin in Omniverse is a virtual space that every engineer and every car in the fleet is connected to,” Huang said.

The ‘Instrument for Your Life’s Work’

Huang wrapped up with four points.

NVIDIA is now a 3-chip company – offering GPUs, CPUs, and DPUs.

NVIDIA is a software platform company and is dedicating enormous investment in NVIDIA AI and NVIDIA Omniverse.

NVIDIA is an AI company with Megatron, Jarvis, Merlin, Maxine, Isaac, Metropolis, Clara, and DRIVE, and pre-trained models you can customize with TAO.

NVIDIA is expanding AI with DGX for researchers, HGX for cloud, EGX for enterprise and 5G edge, and AGX for robotics.

“Mostly,” Huang said. “NVIDIA is the instrument for your life’s work.”

It’s time to hail the new era of transportation.

During his keynote at the GPU Technology Conference today, NVIDIA founder and CEO Jensen Huang outlined the broad ecosystem of companies developing next-generation robotaxis on NVIDIA DRIVE. These forward-looking manufacturers are set to transform the way we move with safer, more efficient vehicles for everyday mobility.

The world moves 2 trillion miles a year. With some of those miles traveled through a mobility service, the opportunity for innovation is tremendous.

Robotaxis are fully autonomous vehicles that can operate without human supervision in geofenced areas, such as cities or residential communities. With a set of high-resolution sensors and a supercomputing platform in place of a driver, they can safely operate 24 hours a day, seven days a week.

Currently, the industry is working to roll out level 2+ AI-assisted driving features in private vehicles, with experts forecasting level 4 vehicles arriving later this decade.

And as a safer alternative to current modes of transit, robotaxis are expected to draw quick adoption once deployed at scale, making up more than 5 percent of vehicle miles traveled worldwide by 2030.

Achieving this mobility revolution requires centralized, high-performance compute. The amount of sensor data a robotaxi needs to process is 100 times greater than today’s most advanced vehicles. The complexity in software also increases exponentially, with an array of redundant and diverse deep neural networks running simultaneously as part of an integrated software stack.

NVIDIA is the only company that enables this level of AI development from end to end, which is why virtually every robotaxi maker and supplier is using its GPU-powered offerings.

Redesigning the Wheel

Some companies are approaching robotaxi development from square one, introducing entirely new vehicles purpose-built for autonomous ride-hailing.

Cruise, the San Francisco self-driving company, unveiled the Cruise Origin robotaxi in early 2020. The all-electric, self-driving, shared vehicle was purpose-built in partnership with GM and Honda to transform what it means to travel and commute in a city.

Cruise leverages the high-performance, energy-efficient compute of NVIDIA DRIVE GPUs to process the massive amounts of data its fleet collects on San Francisco’s chaotic streets in real time. The result is a safer, cleaner and more efficient transportation alternative for city dwellers.

“In a single day we ingest and process what would be equivalent to multiple Netflix libraries,” said Mo ElShenawy, senior vice president of engineering at Cruise during a GTC session. “With NVIDIA DRIVE GPUs, we’re able to use this data to catch our robotaxis up with human evolution.”

In December, robotaxi maker Zoox took the wraps off its rider-focused autonomous vehicle. It features four-wheel steering, allowing it to pull into tight curb spaces without parallel parking. The vehicle is also bidirectional, so there is no fixed front or back end. It can pull forward into a driveway and forward out onto the road without reversing. In the case of an unexpected road closure, the vehicle can simply flip directions or use four-wheel steering to turn around. No reversing required.

Inside the vehicle, carriage seating facilitates clear visibility of the vehicle’s surroundings as well as socializing. Each seat has the same amount of space and delivers the same experience — there’s no bad seat in the house. Carriage seating also makes room for a wider aisle, allowing passengers to easily pass by each other without getting up or contorting into awkward positions.

Zoox was able to optimize this robotaxi design with centralized, high-performance compute built on NVIDIA DRIVE.

“Working with NVIDIA, that’s allowed us to get a couple of orders more magnitude of computation done with the same amount of power, just over the last decade, and that makes a lot of difference,” said Zoox CTO Jesse Levinson during a GTC session.

Moving the World

These companies are also delivering autonomous innovation worldwide.

This past year, self-driving startup AutoX launched a commercial autonomous ride-hailing service in Shenzhen, China.

The 100-vehicle AutoX fleet uses the NVIDIA DRIVE platform for AI compute, achieving up to 2,000 trillion operations per second to power the numerous redundant and deep neural networks for full self-driving.

Based in the United Kingdom, Oxbotica has developed a “Universal Autonomy” platform that enables companies to build a variety of autonomous vehicles, including robotaxis. Its upcoming Selenium platform leverages NVIDIA DRIVE Orin to achieve level 4 self-driving capabilities.

Additionally, ride-hailing giant Didi Chuxing is developing level 4 autonomous vehicles for its mobility services using NVIDIA DRIVE and AI technology. Delivering 10 billion passenger trips per year, DiDi is working toward the safe, large-scale application of autonomous driving technology.

NVIDIA Inception member Pony.AI is collaborating with global automakers such as Toyota and Hyundai, developing a robotaxi fleet with the NVIDIA DRIVE AGX platform at its core.

With the power of NVIDIA DRIVE, these robotaxi companies can continue to roll out this transformative technology to more consumers, ushering in a new age in mobility.

Enterprises rely on GPU virtualization to keep their workforces productive, wherever they work. And NVIDIA virtual GPU (vGPU) performance has become essential to powering a wide range of graphics- and compute-intensive workloads from the cloud and data center.

Now, designers, engineers and knowledge workers across industries can experience accelerated performance with the NVIDIA A10 and A16 GPUs.

Based on the NVIDIA Ampere architecture, A10 and A16 deliver more power, memory and user density to boost any workflow, from graphics and AI to VDI. And when combined with NVIDIA vGPU software, the new GPUs greatly improve user experience, performance and flexibility.

A10 Delivers Powerful, Flexible Virtual Workstation Performance

Professionals are increasingly using advanced technologies like real-time ray tracing, AI, compute, simulation and virtual reality for their work. But to run these workflows, and with employee mobility crucial today, they require more power and flexibility to work from anywhere.

The NVIDIA A10 combined with NVIDIA RTX Virtual Workstation software delivers the performance to efficiently power these complex workflows while ensuring employees get the best user experience possible.

With virtual workstations powered by the A10, businesses can deliver enhanced graphics and video with AI-accelerated applications from mainstream enterprise servers.

Since the A10 can support graphics and AI workloads on virtualized infrastructure, data center administrators can flexibly provision resources and take advantage of any underutilized compute power to run AI inference or VDI workloads.

The A10 combines second-generation RT Cores and third-generation Tensor Cores to enrich graphics and video applications with powerful AI. It’s built specifically for graphics, media and game developer workstations, delivering 2.5x faster graphics performance and over 2.5x the inference performance compared to the previous generation NVIDIA T4 Tensor Core GPU.

Users can also run inference workloads on the A10 with NVIDIA AI Enterprise software and achieve bare-metal performance. The A10 includes new streaming microprocessors with 24GB of GDDR6 memory, enabling versatile graphics, rendering, AI and compute performance. The single-wide, full-height, full-length PCIe form factor enables GPU server density, often five to six GPUs per server.

A16 Enhances VDI User Experience for Remote Workers

With the rising adoption of web conferencing and video collaboration tools, the remote work environment is here to stay. According to an IDC survey, 87 percent of U.S. enterprises expect their employees to continue working from home three or more days per week once mandatory pandemic closures are lifted.⁽¹⁾

Knowledge workers use multiple devices and monitors to efficiently do their work. They also require easy access to productivity tools and applications and need to collaborate with remote teammates. Everything from email and web browsing to video conferencing and streaming can benefit from GPU acceleration — and NVIDIA A16 provides that powerful performance by delivering the next generation of VDI.

The A16 combined with NVIDIA vPC software is ideal for providing graphics-rich VDI and an enhanced user experience for knowledge workers. It offers improved user density versus the previous generation M10, with up to 64 concurrent users per board and reduces the total cost of ownership by up to 20 percent.

Virtual desktops powered by NVIDIA vPC software and the A16 deliver an experience indistinguishable from a physical PC, which allows remote workers to seamlessly transition between working at the office and at home.

GPU-accelerated VDI with A16 and NVIDIA vPC also provides increased frame rates and lower end-user latency, so productivity applications and tools are more responsive, and remote workers achieve the optimal user experience.

Availability

NVIDIA A10 is supported as part of NVIDIA-Certified Systems, in the on-prem data center, in the cloud and at the edge, and will be available starting this month. Learn more about the NVIDIA A10 by watching a replay of NVIDIA CEO Jensen Huang’s GTC keynote address.

NVIDIA A16 will be available later this year.

_{(1) IDC Press Release, Mobile Workers Will Be 60% of the Total U.S. Workforce by 2024, According to IDC, September 2020}

Volvo Cars is extending its long-held legacy of safety far into the future.

The global automaker announced during the GTC keynote today that it will use NVIDIA DRIVE Orin to power the autonomous driving computer in its next-generation cars. The decision deepens the companies’ collaboration to even more software-defined model lineups, beginning with the next-generation XC90, set to debut next year.

Volvo Cars has been using the high-performance, energy-efficient compute of NVIDIA DRIVE since 2016 and developing AI-assisted driving features for new models on NVIDIA DRIVE Xavier with software developed in-house and by Zenseact, Volvo Cars’ autonomous driving software development company.

Building safe self-driving cars is one of the most complex computing challenges today. Advanced sensors surrounding the car generate enormous amounts of data that must be processed in a fraction of a second. That’s why NVIDIA developed Orin, the industry’s most advanced, functionally safe and secure, software-defined autonomous vehicle computing platform.

Orin is software compatible with Xavier, allowing customers to leverage their existing development investments. It’s also scalable — with a range of configurations and even able to deliver unsupervised driverless operation.

Volvo Cars’ next-generation vehicle architecture will be hardware-ready for autonomous driving from production start. Its unsupervised autonomous driving feature, called Highway Pilot, will be activated when it’s verified to be safe for individual geographic locations and conditions.

Redundancy and Diversity for Any Adversity

The NVIDIA DRIVE platform is architected for redundancy and diversity to deliver the highest level of safety.

Like its predecessors, NVIDIA Orin (pictured below) maintains this safety architecture with the highest possible compute performance. The system-on-a-chip (SoC) achieves up to 254 TOPS and is designed to handle the large number of applications and deep neural networks that run simultaneously in autonomous vehicles and robots, while achieving systematic safety standards such as ISO 26262 ASIL-D.

By combining the compute performance of Orin with software developed in-house and by Zenseact, and state-of-the-art sensors such as LiDAR and radar, Volvo Cars’ upcoming generations of intelligent cars will feature safe and robust AI capabilities.

Continuous Improvement

The next generation of vehicles will be state-of-the-art data centers on wheels. They’ll be richly programmable and receive software updates over the air.

These software-defined capabilities will deliver new skills and features that will delight drivers and passengers for the life of the car.

By centralizing the vehicle’s compute on NVIDIA DRIVE Orin, Volvo Cars’ next generation cars will be safer, more personal and more sustainable, that become better and smarter every day. Even when these cars aren’t in autonomous driving mode, they can still improve the safety of their occupants by anticipating and reacting to hazards faster than a human driver.

With an architecturally coherent and programmable fleet, Volvo Cars will extend its legacy of safety and quality far into the future, nurturing a growing installed base with its upcoming cars to offer software upgradeable applications for the entire life of the car.