Monthly Archives: March 2024

Designing, simulating and bringing up modern data centers is incredibly complex, involving multiple considerations like performance, energy efficiency and scalability.

It also requires bringing together a team of highly skilled engineers across compute and network design, computer-aided design (CAD) modeling, and mechanical, electrical and thermal design.

NVIDIA builds the world’s most advanced AI supercomputers and at GTC unveiled its latest — a large cluster based on the NVIDIA GB200 NVL72 liquid-cooled system. It consists of two racks, each containing 18 NVIDIA Grace CPUs and 36 NVIDIA Blackwell GPUs, connected by fourth-generation NVIDIA NVLink switches.

On the show floor, NVIDIA demoed this fully operational data center as a digital twin in NVIDIA Omniverse, a platform for connecting and building generative AI-enabled 3D pipelines, tools, applications and services.

To bring up new data centers as fast as possible, NVIDIA first built its digital twin with software tools connected by Omniverse. Engineers unified and visualized multiple CAD datasets in full physical accuracy and photorealism in Universal Scene Description (OpenUSD) using the Cadence Reality digital twin platform, powered by NVIDIA Omniverse APIs.

Design, Simulate and Optimize With Enhanced Efficiency and Accuracy

The new GB200 cluster is replacing an existing cluster in one of NVIDIA’s legacy data centers. To start the digital build-out, technology company Kinetic Vision scanned the facility using the NavVis VLX wearable lidar scanner to produce highly accurate point cloud data and panorama photos.

Then, Prevu3D software was used to remove the existing clusters and convert the point cloud to a 3D mesh. This provided a physically accurate 3D model of the facility, in which the new digital data center could be simulated.

Engineers combined and visualized multiple CAD datasets with enhanced precision and realism by using the Cadence Reality platform. The platform’s integration with Omniverse provided a powerful computing platform that enabled teams to develop OpenUSD-based 3D tools, workflows and applications.

Omniverse Cloud APIs also added interoperability with more tools, including PATCH MANAGER and NVIDIA Air. With PATCH MANAGER, the team designed the physical layout of their cluster and networking infrastructure, ensuring that cabling lengths were accurate and routing was properly configured.

The team used Cadence’s Reality Digital Twin solvers, accelerated by NVIDIA Modulus APIs and NVIDIA Grace Hopper, to simulate airflows, as well as the performance of the new liquid-cooling systems from partners like Vertiv and Schneider Electric. The integrated cooling systems in the GB200 trays were simulated and optimized using solutions from Ansys, which brought simulation data into the digital twin.

The demo showed how digital twins can allow users to fully test, optimize and validate data center designs before ever producing a physical system. By visualizing the performance of the data center in the digital twin, teams can better optimize their designs and plan for what-if scenarios.

Users can also enhance data center and cluster designs by balancing disparate sets of boundary conditions, such as cabling lengths, power, cooling and space, in an integrated manner — enabling engineers and design teams to bring clusters online much faster and with more efficiency and optimization than before.

Learn more about NVIDIA Omniverse and NVIDIA Blackwell.

A sharp increase in generative AI deployments is driving business innovation for enterprises across industries. But it’s also posing significant challenges for their IT teams, as slowdowns from long and complex infrastructure deployment cycles prevent them from quickly spinning up AI workloads using their own data.

To help overcome these barriers, NVIDIA has introduced a storage partner validation program for NVIDIA OVX computing systems. The high-performance storage systems leading the way in completing the NVIDIA OVX storage validation are DDN, Dell PowerScale, NetApp, Pure Storage and WEKA.

NVIDIA OVX servers combine high-performance, GPU-accelerated compute with high-speed storage access and low-latency networking to address a range of complex AI and graphics-intensive workloads. Chatbots, summarization and search tools, for example, require large amounts of data, and high-performance storage is critical to maximize system throughput.

To help enterprises pair the right storage with NVIDIA-Certified OVX servers, the new program provides a standardized process for partners to validate their storage appliances. They can use the same framework and testing that’s needed to validate storage for the NVIDIA DGX BasePOD reference architecture.

To achieve validation, partners must complete a suite of NVIDIA tests measuring storage performance and input/out scaling across multiple parameters that represent the demanding requirements of various enterprise AI workloads. This includes combinations of different I/O sizes, varying numbers of threads, buffered I/O vs. direct I/O, random reads, re-reads and more.

Each test is run multiple times to verify the results and gather the required data, which is then audited by NVIDIA engineering teams to determine whether the storage system has passed.

The program offers prescriptive guidance to ensure optimal storage performance and scalability for enterprise AI workloads with NVIDIA OVX systems. But the overall design remains flexible, so customers can tailor their system and storage choices to fit their existing data center environments and bring accelerated computing to wherever their data resides.

Generative AI use cases have fundamentally different requirements than traditional enterprise applications, so IT teams must carefully consider their compute, networking, storage and software choices to ensure high performance and scalability.

NVIDIA-Certified Systems are tested and validated to provide enterprise-grade performance, manageability, security and scalability for AI workloads. Their flexible reference architectures help deliver faster, more efficient and more cost-effective deployments than independently building from the ground up.

Powered by NVIDIA L40S GPUs, OVX servers include NVIDIA AI Enterprise software with NVIDIA Quantum-2 InfiniBand or NVIDIA Spectrum-X Ethernet networking, as well as NVIDIA BlueField-3 DPUs. They’re optimized for generative AI workloads, including training for smaller LLMs (for example, Llama 2 7B or 70B), fine-tuning existing models and inference with high throughput and low latency.

NVIDIA-Certified OVX servers are now available and shipping from global system vendors, including GIGABYTE, Hewlett Packard Enterprise and Lenovo. Comprehensive, enterprise-grade support for these servers is provided by each system builder, in collaboration with NVIDIA.

Availability 

Validated storage solutions for NVIDIA-Certified OVX servers are now available, and reference architectures will be published over the coming weeks by each of the storage and system vendors. Learn more about NVIDIA OVX Systems.

The latest advances in quantum computing include investigating molecules, deploying giant supercomputers and building the quantum workforce with a new academic program.

Researchers in Canada and the U.S. used a large language model to simplify quantum simulations that help scientists explore molecules.

“This new quantum algorithm opens the avenue to a new way of combining quantum algorithms with machine learning,” said Alan Aspuru-Guzik, a professor of chemistry and computer science at the University of Toronto, who led the team.

The effort used CUDA-Q, a hybrid programming model for GPUs, CPUs and the QPUs quantum systems use. The team ran its research on Eos, NVIDIA’s H100 GPU supercomputer.

Software from the effort will be made available for researchers in fields like healthcare and chemistry. Aspuru-Guzik will detail the work in a talk at GTC.

Quantum Scales for Fraud Detection

At HSBC, one of the world’s largest banks, researchers designed a quantum machine learning application that can detect fraud in digital payments.

The bank’s quantum machine learning algorithm simulated a whopping 165 qubits on NVIDIA GPUs. Research papers typically don’t extend beyond 40 of these fundamental calculating units quantum systems use.

HSBC used machine learning techniques implemented with CUDA-Q and cuTensorNet software on NVIDIA GPUs to overcome challenges simulating quantum circuits at scale. Mekena Metcalf, a quantum computing research scientist at HSBC (pictured above), will present her work in a session at GTC.

Raising a Quantum Generation

In education, NVIDIA is working with nearly two dozen universities to prepare the next generation of computer scientists for the quantum era. The collaboration will design curricula and teaching materials around CUDA-Q.

“Bridging the divide between traditional computers and quantum systems is essential to the future of computing,” said Theresa Mayer, vice president for research at Carnegie Mellon University. “NVIDIA is partnering with institutions of higher education, Carnegie Mellon included, to help students and researchers navigate and excel in this emerging hybrid environment.”

To help working developers get hands-on with the latest tools, NVIDIA co-sponsored QHack, a quantum hackathon in February. The winning project, developed by Gopesh Dahale of Qkrishi — a quantum company in Gurgaon, India — used CUDA-Q to develop an algorithm to simulate a material critical in designing better batteries.

A Trio of New Systems

Two new systems being deployed further expand the ecosystem for hybrid quantum-classical computing.

The largest of the two, ABCI-Q at Japan’s National Institute of Advanced Industrial Science and Technology, will be one of the largest supercomputers dedicated to research in quantum computing. It will use CUDA-Q on NVIDIA H100 GPUs to advance the nation’s efforts in the field.

In Denmark, the Novo Nordisk Foundation will deploy an NVIDIA DGX SuperPOD, half of which will be dedicated to research in quantum computing as part of the country’s national plan to advance the technology.

The new systems join Australia’s Pawsey Supercomputing Research Centre, which announced in February it will run CUDA-Q on NVIDIA Grace Hopper Superchips at its National Supercomputing and Quantum Computing Innovation Hub.

Partners Drive CUDA Quantum Forward

In other news, Israeli startup Classiq released at GTC a new integration with CUDA-Q. Classiq’s quantum circuit synthesis lets high-level functional models automatically generate optimized quantum programs, so researchers can get the most out of today’s quantum hardware and expand the scale of their work on future algorithms.

Software and service provider QC Ware is integrating its Promethium quantum chemistry package with the just-announced NVIDIA Quantum Cloud.

ORCA Computing, a quantum systems developer headquartered in London, released results running quantum machine learning on its photonics processor with CUDA-Q. In addition, ORCA was selected to build and supply a quantum computing testbed for the UK’s National Quantum Computing Centre which will include an NVIDIA GPU cluster using CUDA-Q.

Nvidia and Infleqtion, a quantum technology leader, partnered to bring cutting-edge quantum-enabled solutions to Europe’s largest cyber-defense exercise with NVIDIA-enabled Superstaq software.

A cloud-based platform for quantum computing, qBraid, is integrating CUDA-Q into its developer environment. And California-based BlueQubit described in a blog how NVIDIA’s quantum technology, used in its research and GPU service, provides the fastest and largest quantum emulations possible on GPUs.

Get the Big Picture at GTC

To learn more, watch a session about how NVIDIA is advancing quantum computing and attend an expert panel on the topic, both at NVIDIA GTC, a global AI conference, running March 18-21 at the San Jose Convention Center.

And get the full view from NVIDIA founder and CEO Jensen Huang in his GTC keynote.

Video conferencing has allowed many to be productive from anywhere.

Now NVIDIA is boosting the productivity of the developers of video conferencing, call center and streaming applications within the $10 billion industry by allowing them to easily integrate AI into their workflows.

The new release of the Maxine AI Developer Platform transforms the creation of state-of-the-art, real-time video conferencing applications with features enabling enhanced user flexibility, engagement and efficiency.

Available through the NVIDIA AI Enterprise software platform, Maxine allows developers to tap into the latest AI-driven features — such as enhanced video and audio quality and augmented reality effects — to turn users’ everyday video calls into engaging, collaborative experiences.

Expanding Video Conferencing With New Maxine Features 

The Maxine AI Developer Platform enables developers to easily access and integrate real-time, AI-enhanced features that increase the quality of engagement for video conferencing users.

Features like noise reduction, video denoising and upscaling, and studio voice improve the quality of audio and video streams. With advanced capabilities like eye-gaze correction, live portrait and future features such as video relighting and cloud microservice Maxine 3D, developers can enhance video conferencing engagement and personal connection.

The platform extends the utility of the state-of-the-art AI models for audio, video and augmented reality effects with multiple ways for developers to deliver Maxine features with offerings of software development kits, microservices, and even application programming interface (API) endpoints delivered from NVIDIA’s cloud infrastructure.

Maxine production feature updates available now include:

• Eye Contact: The improved eye contact model provides gaze redirection with natural eye movements for deeper meeting participant engagement.
• Voice Font: This new model matches the speaker’s voice to a target voice while keeping linguistic information and prosody (rhythm and tone) unchanged.
• Background Noise Reduction (BNR) 2.0: This model updates noise reduction for human listening and for language encoding with a specific effort to decrease encoding word error rates.

New features available for early access this spring include:

• Speech Live Portrait: This model allows a user to drive their portrait with direct speech or any audio source, allowing users to always look their best during a conference call.
• Studio Voice: This model enables ordinary headset, laptop and desktop microphones to deliver the sound of a high-end studio mic, allowing users to always sound their best during a conference call.

The Maxine early access program shares preproduction and prerelease builds of upcoming features in order to get feedback from developers on the utility and refinement of Maxine models. In this release we are asking developers for feedback on features early in the development pipeline including:

• Maxine 3D: Previously shown as a research demonstration at SIGGRAPH 2023, this cloud microservice offers a new level of engagement for video conferencing with real-time NeRF technology lifting 2D video to 3D.
• Video Relighting: This new model uses a high-dynamic-range image to light the user, enabling seamless matching of user lighting with various background images.
• API Endpoints: API Endpoints offer developers the flexibility of accessing Maxine features through NVIDIA cloud infrastructure, making Maxine integration even easier.

Jugo and Arsenal Football Club Score Major Goals 

Sporting events are the ultimate human experience, uniting teams and fans beyond borders and language barriers. Jugo, using Maxine’s AI Green Screen feature, offers a digital platform for virtual events that enables companies to create immersive experiences with Unreal Engine that bring fans together from all over the world without the use of a full production studio.

Arsenal FC, a powerhouse franchise in England’s Premier League, is collaborating with Jugo to revolutionize the way the soccer club engages with its 600 million global fan base. The collaboration offers new virtual sports entertainment experiences to boost engagement for global supporters. Jugo brings the power of real, human interaction into Arsenal events, creating realistic virtual connections between supporters and the club’s sports heroes.

“The Jugo Experience platform is transforming the market for brands in their pursuit of global awareness and engagement,” said Richard Stirk, CEO of Jugo Experience. “Arsenal F.C. is the perfect example of a global brand extension. The flexibility in creating an immersive brand experience is a key to Jugo’s offering and the Maxine AI Developer Platform is a basic building block of this flexibility.”

Setting a New Standard of AI-Enhanced Video Conferencing 

Among the first customers to tap into the newest set of features within the early access program to create a professional audio-visual studio from commodity cameras and microphones are Gemelo, Pexip, Spectacle and VideoRequest.

“Gemelo has been involved in testing prerelease builds of Maxine models for a number of years now, and we value the chance to give early input on Maxine features as they’re developed,” said Paul Jaski, CEO of Gemelo. “The latest feature, Speech Live Portrait, will provide our customers with greater flexibility in creating customized video messaging, opening the doors to a new era of personalization.”

“Pexip welcomes the chance to test development versions of Maxine features and help guide the final product models,” said Ian Mortimer, chief technology officer at Pexip. “In testing the newest version of Maxine BNR, we are seeing significant improvements in intelligibility and speech quality and plan to continue refining our testing parameters to help optimize for accuracy in AI translation pipelines.”

“The NVIDIA Maxine Eye Contact API significantly simplified our path to providing engaging video processing capabilities to the users of our Spectacle app, eliminating the need to worry about infrastructure and resource-intensive integrations,” said Benjamin Portman, president of Spectacle. “With it, we were able to create a proof of concept within a matter of days, speeding up our production application deployment timeline.”

“Our early testing of Maxine Studio Voice enabled an impressive look into what is now possible with AI-enhanced production and video testimonials,” said Joe Tyler, chief technology officer at VideoRequest. “The new Maxine BNR and Eye Contact features will help elevate the quality of our customer’s videos by overcoming their challenging recording environments.”

Availability 

Learn more about NVIDIA Maxine, which is available now on NVIDIA AI Enterprise.

See notice regarding software product information.

NVIDIA Edify, a multimodal architecture for visual generative AI, is entering a new dimension.

3D asset generation is among the latest capabilities Edify offers developers and visual content providers, who will also be able to exert more creative control over AI image generation.

Multimedia content and data provider Shutterstock is rolling out early access to an API, or application programming interface, built on the Edify architecture that lets creators use text prompts or images to rapidly generate 3D objects for virtual scenes.

Visual content creator and marketplace Getty Images will add custom fine-tuning capabilities to its commercially safe Generative AI service, helping enterprise customers generate visuals that adhere to brand guidelines and styles. The service will also incorporate new features to offer customers even further control of their generated images.

Developers can test drive pretrained Edify models by Getty Images and Shutterstock as APIs through NVIDIA NIM, a collection of microservices for inference announced today at NVIDIA GTC. Developers can also train and deploy their own generative AI models using the Edify architecture through NVIDIA Picasso, an AI foundry built on NVIDIA DGX Cloud.

NVIDIA and Adobe are collaborating to bring new 3D generative AI technologies built on Edify to millions of Firefly and Creative Cloud creators.

Livestreaming platform Be.Live is using the NVIDIA Picasso foundry service to provide real-time generative AI that enables the automated creation of visuals and an interactive experience for audiences. Bria, a holistic platform tailored for businesses developing responsible visual generative AI, has adopted Picasso to run inference. And creative studio Cuebric is enhancing filmmaking and content creation by developing Picasso-powered generative AI applications to build immersive virtual environments.

Speedy 3D: Shutterstock 3D AI Generator Now in Early Access

Shutterstock’s 3D AI Services, available in early access, will enable creators to generate virtual objects for set dressing and ideation. This capability can drastically reduce the time needed to prototype a scene, giving artists more time to focus on hero characters and objects.

Shutterstock 3D generator in action. Video courtesy of Shutterstock.

Using the tools, creative professionals will be able to rapidly create assets from text prompts or reference images and choose from a selection of popular 3D formats to export their files. The Edify 3D-based service will also come with built-in safeguards to filter generated content.

The commercially safe model was trained on Shutterstock’s licensed data. Shutterstock has compensated hundreds of thousands of artists, with anticipated payments to millions more, for the role their content IP has played in training generative technology.

At GTC, HP and Shutterstock are showcasing a collaboration to enhance custom 3D printing using Edify 3D, providing designers with limitless prototype options.

Shutterstock’s 3D AI generator enables designers to rapidly iterate on concepts, creating digital assets that HP can convert to 3D printable models through automated workflows. HP 3D printers will then turn these models into physical prototypes to help inspire product designs.

Mattel is enabling 3D generative AI capabilities from Shutterstock that can accelerate the design ideation process. With AI tools, toy designers can visualize their ideas for new products with simple text descriptions. By lowering the technical barrier to creating high-fidelity concept design, designers can explore a broader pool of their ideas and iterate faster.

Shutterstock is also building Edify-based tools to light 3D scenes using 360 HDRi environments generated from text or image prompts.

Dassault Systèmes, through its leading 3DEXCITE applications for 3D content creation, and CGI studio Katana are incorporating Shutterstock generative 360 HDRi APIs into their workflows based on NVIDIA Omniverse, a computing platform for developing Universal Scene Description (OpenUSD)-based 3D workflows and applications.

Accenture Song, the world’s largest tech-powered creative group, is using the Omniverse platform to generate high-fidelity Defender vehicles from computer-aided design data for marketing purposes. Coupled with generative AI microservices powered by Edify, Accenture Song is enabling the creation of cinematic, interactive 3D environments via conversational prompts. The result is a fully immersive 3D scene that harmonizes realistic generated environments with a digital twin of the Defender vehicle.

Take Control: Turn Creative Vision Into Reality With Custom Fine-Tuning From Getty Images

Getty Images continues to expand the capabilities offered through its commercially safe generative AI service, which provides users indemnification for the content they generate.

At January’s CES show, Getty Images released Edify-powered APIs for inpainting, to add, remove or replace objects in an image, and outpainting, to expand the creative canvas. Those features are now available on both Gettyimages.com and iStock.com.

Starting in May, Getty Images will also offer services to custom fine-tune the Edify foundation model to a company’s brand and visual style.

The services will feature a no-code, self-service method for companies to upload a proprietary dataset, review automatically generated tags, submit fine-tuning tasks and review the results before deploying to production.

As part of custom fine-tuning tools, Getty Images will release a collection of APIs that provide finer control over image output, one of the biggest challenges in generative AI.

Developers will soon be able to access Sketch, Depth and Segmentation features — which allow users to provide a sketch to guide the AI’s image generation; copy the composition of a reference image via depth map; and segment parts of an image to add, remove or retouch a character or object.

Getty Images’ API services are already being used by leading creatives and advertisers, including:

• Dentsu Inc.: The Japan-based ad agency is using Getty Images’ generative AI API service to power MAFA: Manga Anime For All, an app that can create manga and anime-style content for marketing use cases. Dentsu Creative is also using NVIDIA Picasso to fine-tune Getty Images’ model for leading membership warehouse retailer Sam’s Club.
• McCann: The creative agency harnessed generative AI to help develop an innovative game for its client Reckitt’s over-the-counter cold medicine Mucinex, in which customers can interact with the brand’s mascot.
• Refik Anadol Studio: Known for using generative AI in its artwork, the studio will be showcasing a rainforest-inspired art installation at GTC, created using Getty Images’ AI model fine-tuned with Refik’s rainforest catalog.
• WPP: The marketing and communications services company is partnering with The Coca-Cola Company to explore how fine-tuning Getty Images’ model can help to build custom visuals that meet brand styles and guidelines.

Learn more about NVIDIA Picasso and try Edify-powered NIMs from Getty Images and Shutterstock at ai.nvidia.com.

Discover the latest in generative AI at NVIDIA GTC, a global AI developer conference, running in San Jose, Calif., and online through Thursday, March 21. 

Watch the GTC keynote address by NVIDIA founder and CEO Jensen Huang below:

Collage at top shows assets created by Edify-powered Shutterstock 3D AI generator on left, courtesy of Shutterstock. Images on right show Edify sketch-to-image capabilities, demonstrated by NVIDIA.

The journey to the future of secure communications is about to jump to warp drive.

NVIDIA cuPQC brings accelerated computing to developers working on cryptography for the age of quantum computing. The cuPQC library harnesses the parallelism of GPUs for their most demanding security algorithms.

Refactoring Security for the Quantum Era  

Researchers have known for years that quantum computers will be able to break the public keys used today to secure communications. As these systems approach readiness, government and industry initiatives have been ramping up to address this vital issue.

The U.S. National Institute of Standards and Technology, for example, is expected to introduce the first standard algorithms for post-quantum cryptography as early as this year.

Cryptographers working on advanced algorithms to replace today’s public keys need powerful systems to design and test their work.

Hopper Delivers up to 500x Speedups With cuPQC

In its first benchmarks, cuPQC accelerated Kyber — an algorithm proposed as a standard for securing quantum-resistant keys — by up to 500x running on an NVIDIA H100 Tensor Core GPU compared with a CPU.

The speedups will be even greater with NVIDIA Blackwell architecture GPUs, given Blackwell’s enhancements for the integer math used in cryptography and other high performance computing workloads.

“Securing data against quantum threats is a critically important problem, and we’re excited to work with NVIDIA to optimize post-quantum cryptography,” said Douglas Stebila, co-founder of the Open Quantum Safe project, a group spearheading work in the emerging field.

Accelerating Community Efforts

The project is a part of the newly formed Post-Quantum Cryptography Alliance, hosted by the Linux Foundation.

The alliance funds open source projects to develop post-quantum libraries and applications. NVIDIA is a member of the alliance with seats on both its steering and technical committees.

NVIDIA is also collaborating with cloud service providers such as Amazon Web Services (AWS), Google Cloud and Microsoft Azure on testing cuPQC.

In addition, leading companies in post-quantum cryptography such as EvolutionQ, PQShield, QuSecure and SandboxAQ are collaborating with NVIDIA, many with plans to integrate cuPQC into their offerings.

“Different use cases will require a range of approaches for optimal acceleration,” said Ben Packman, a senior vice president at PQShield. “We are delighted to explore cuPQC with NVIDIA.”

Learn More at GTC

Developers working on post-quantum cryptography can sign up for updates on cuPQC here.

To learn more, watch a session about how NVIDIA is advancing quantum computing and attend an expert panel on the topic at NVIDIA GTC, a global AI conference, running through March 21 at the San Jose Convention Center and online.

Get the full view from NVIDIA founder and CEO Jensen Huang in his GTC keynote.

Generative AI is capable of creating more realistic verbal and facial expressions for digital humans than ever before.

This week at GDC 2024, NVIDIA announced that leading AI application developers across a wide range of industries are using NVIDIA digital human technologies to create lifelike avatars for commercial applications and dynamic game characters. NVIDIA enables developers with state-of-the-art digital human technologies, including NVIDIA ACE for speech and animation, NVIDIA NeMo for language, and NVIDIA RTX for ray-traced rendering.

Developers showcased new digital human technology demos that used NVIDIA ACE microservices at GDC.

Embracing ACE: Partners Transforming Pixels Into Personalities  

Top game and digital human developers are pioneering ways ACE and generative AI technologies can be used to transform interactions between players and NPCs in games and applications.

Developers embracing ACE include: Convai, Cyber Agent, Data Monsters, Deloitte, HippocraticAI, IGOODI, Inworld AI, Media.Monks, miHoYo, NetEase Games, Perfect World Games, Openstream, OurPalm, Quantiphi, Rakuten Securities, Slalom, SoftServe, Tencent, Ubisoft, UneeQ and Unions Avatars.

Demos Showcase New NVIDIA Digital Human Technologies

NVIDIA worked with developers Inworld AI and UneeQ on a series of new demos to display the potential of digital human technologies.

Inworld AI created Covert Protocol in partnership with NVIDIA, allowing players to become a skilled private detective, pushing the possibilities of non-playable character interactions. The demo taps into NVIDIA Riva automatic speech recognition (ASR) and NVIDIA Audio2Face microservices alongside the Inworld Engine.

The Inworld Engine brings together cognition, perception and behavior systems to create an immersive narrative along with the beautifully crafted RTX-rendered environments and art.

UneeQ is a digital human platform specialized in creating high-fidelity AI-powered 3D avatars for a range of enterprise applications. UneeQ’s digital humans power interactive experiences for brands enabling them to communicate with customers in real-time to give them confidence in their purchases. UneeQ integrated NVIDIA Audio2Face microservice into its platform and combined it with Synanim ML to create highly realistic avatars for a better customer experience and engagement.

New NVIDIA RTX Technologies for Dynamic Scenes

NVIDIA RTX revolutionized gaming several years ago by offering a collection of rendering technologies that enable real-time path tracing in games and applications.

The latest addition, Neural Radiance Cache (NRC), is an AI-driven RTX algorithm to handle indirect lighting in fully dynamic scenes, without the need to bake static lighting for geometry and materials beforehand.

Adding flexibility for developers, NVIDIA is introducing Spatial Hash Radiance Cache (SHaRC), which offers similar benefits as NRC but without using a neural network, and with compatibility on any DirectX or Vulkan ray tracing-capable GPU.

RTX. It’s On: More RTX and DLSS 3.5 Titles 

There are now over 500 RTX games and applications that have revolutionized the ways people play and create with ray tracing, NVIDIA DLSS and AI-powered technologies.  And gamers have a lot to look forward to with more full ray tracing and DLSS 3.5 titles coming.

Our latest innovation, NVIDIA DLSS 3.5, features new DLSS Ray Reconstruction technology. When activated, DLSS Ray Reconstruction replaces hand-tuned ray tracing denoisers with a new unified AI model that enhances ray tracing in supported games, elevating image quality to new heights.

Full Ray Tracing and DLSS 3.5 are coming to both Black Myth: Wukong and NARAKA: BLADEPOINT. And Portal with RTX is available now with RTX DLSS 3.5, enhancing its already beautiful full ray tracing. DLSS 3.5 With Ray Reconstruction will also be coming soon to the NVIDIA RTX Remix Open Beta, enabling modders to add RTX technologies like full ray tracing and DLSS 3.5 into classic games.

Star Wars Outlaws will launch with DLSS 3 and ray-traced effects. Ray tracing joins DLSS 3 in Diablo IV March 26. The First Berserker: Khazan will launch with DLSS 3. And Sengoku Destiny introduced support for DLSS 3 and is available now.

See our Partner Ecosystem at GDC

NVIDIA and our partners will showcase the latest in digital human technologies throughout the week of GDC. Here’s a quick snapshot:

• Inworld AI (Booth P1615): Attendees will get the chance to try out Covert Protocol for themselves live at GDC.
• Oracle Cloud Infrastructure (Booth S941): See Covert Protocol in action, and discover the “code assist” ability of Retrieval Augmented Generation (RAG). Register and join an exclusive networking event with Oracle and NVIDIA AI experts on March 21, open to women and allies in the gaming industry to build connections with leading voices in AI.
• Dell Technologies and International Game Developer Association (Booth S1341): Playtest while building a game-ready asset on a workstation with large GPU memory. Speak to Sophie, an AI-powered assistant created by UneeQ and powered by NVIDIA ACE. Attendees can see the latest debugging and profiling tools for making ray-traced games​ in the latest Nsight Graphics ray tracing demo.
• AWS: Developers can register and join NVIDIA, AWS, game studios, and technology partners as they discuss the game tech used to build, innovate, and maximize growth of today’s games at the AWS for Games Partner Showcase on March 20th.

Stop by these key sessions:

• Transforming Gameplay with AI NPCs: This session featuring Nathan Yu, director of product at Inworld AI, Rajiv Gandhi, master principal cloud architect at Oracle Cloud and Yasmina Benkhoui, generative AI strategic partnerships lead at NVIDIA, will showcase successful examples of developers using AI NPCs to drive core game loops and mechanics that keep players engaged and immersed. Attendees will gain a deeper understanding of the potential of AI NPCs to create new and immersive experiences for players.
• Alan Wake 2: A Deep Dive into Path Tracing Technology: This session will take a deep dive into the path tracing and NVIDIA DLSS Ray Reconstruction technologies implemented in Remedy Entertainment’s Alan Wake 2. Developers can discover how these cutting-edge techniques can enhance the visual experience of games.

Download our show guide to keep this summary on hand while at the show.

Get Started

Developers can start their journey on NVIDIA ACE by applying for our early access program to get in-development AI models.

If you want to explore available models, evaluate and access NVIDIA NIM, a set of easy-to-use microservices designed to accelerate the deployment of generative AI, for RIVA ASR and Audio2Face on ai.nvidia.com today.

RTXGI’s NRC and SHaRC algorithms are also available now as an experimental branch.

AI — already used to connect, analyze and offer predictions based on operating room data — will be critical to the future of surgery, boosting operating room efficiency and clinical decision-making.

That’s why NVIDIA is working with Johnson & Johnson MedTech to test new AI capabilities for the company’s connected digital ecosystem for surgery. It aims to enable open innovation and accelerate the delivery of real-time insights at scale to support medical professionals before, during and after procedures.

J&J MedTech is in 80% of the world’s operating rooms and trains more than 140,000 healthcare professionals each year through its education programs.

Seeking to bring together its legacy and digital ecosystem in surgery with NVIDIA’s leading AI solutions — including the NVIDIA IGX edge computing platform and NVIDIA Holoscan edge AI platform for building medical devices — J&J MedTech can accelerate the infrastructure needed to deploy AI-powered software applications for surgery. IGX and Holoscan can support secure, real-time processing from devices across the operating room to provide clinical insights and improve surgical outcomes.

Unveiled at NVIDIA GTC, the global AI conference taking place March 18-21 in San Jose, Calif., and online, this work could also facilitate the deployment of third-party models and applications developed across the digital surgery ecosystem by providing a common AI compute platform.

“AI models are currently being created by experts in surgery in various parts of the world,” said Shan Jegatheeswaran, vice president and global head of digital at J&J MedTech. “If we can create a trusted, open ecosystem that enables and accelerates coordination, it would create a flywheel of innovation where different groups can collaborate and connect at scale, improving access to advanced analytics across the surgical experience.”

An Open Ecosystem for AI Innovation: Building on NVIDIA Holoscan and IGX 

J&J MedTech is working with NVIDIA to test how industrial-grade edge AI capabilities purpose-built for medical environments could benefit surgery.

“Our connected digital ecosystem will help break down the traditional barriers to entry for developers seeking to build applications and deploy analytics in the operating room,” Jegatheeswaran said. “We’re making it simpler for those who want to participate in the surgical workflow by eliminating the heavy lifting of building a secure, enterprise-grade platform.”

NVIDIA Holoscan accelerates the development and deployment of real-time AI applications to process data streams.

Holoscan includes reference pipelines to build AI applications for a variety of medical use cases, including endoscopy, ultrasound and other sensors. It runs on NVIDIA IGX, which includes NVIDIA Jetson Orin modules, NVIDIA RTX A6000 GPUs and NVIDIA ConnectX networking technology to enable high-speed data streaming from medical devices or operating room video feeds.

NVIDIA supports the IGX software stack with NVIDIA AI Enterprise, the enterprise operating system for production-grade AI.

Fueling Surgical AI With Device Data

The J&J MedTech team envisions the potential of NVIDIA-accelerated edge analytics behind its connected digital ecosystem as an enabler of AI-powered applications fueled by device, patient and other surgical data.

Developers could leverage continuous learning, where an algorithm improves based on data collected by the deployed device. Real-world footage collected by an endoscope, for example, could be used to refine an AI model that identifies organs, tissue and potential tumors in real time on an operating room display to support clinical decision-making.

“Surgical technologies will get more intelligent over time, bringing the power of advanced analytics to surgeons and hospitals,” said Jegatheeswaran. “A collection of AI models could act like driver-assistance technology for surgeons, amplifying their ability to deliver care while reducing cognitive load.”

One example is AI that removes personally identifiable information from surgical videos so they can be used downstream for research purposes — or, when processed in real time, enable hospitals to bring in external experts through telepresence to consult during a surgery while maintaining patient privacy.

Future applications could enable surgeons to interact with chatbots to gain insights about a patient’s medical history or best practices for handling certain complications. Other models could improve operating room efficiency by using video feeds to understand when a procedure is almost complete, alerting the next surgical team that a room will soon be available.

Discover the latest in AI and healthcare at GTC, running in San Jose, Calif., and online through Thursday, March 21. Tune in to a special address on generative AI in healthcare delivered by Kimberly Powell, vice president of healthcare at NVIDIA, on Tuesday at 8 a.m. PT.

Watch the GTC keynote address by NVIDIA founder and CEO Jensen Huang below:

Moving fast to accelerate its AI journey, BNY Mellon, a global financial services company celebrating its 240th anniversary, revealed Monday that it has become the first major bank to deploy an NVIDIA DGX SuperPOD with DGX H100 systems.

Thanks to the strong collaborative relationship between NVIDIA Professional Services and BNY Mellon, the team was able to install and configure the DGX SuperPOD ahead of typical timelines.

The system, equipped with dozens of NVIDIA DGX systems and NVIDIA InfiniBand networking and based on the DGX SuperPOD reference architecture, delivers computer processing performance that hasn’t been seen before at BNY Mellon.

“Key to our technology strategy is empowering our clients through scalable, trusted platforms and solutions,” said BNY Mellon Chief Information Officer Bridget Engle. “By deploying NVIDIA’s AI supercomputer, we can accelerate our processing capacity to innovate and launch AI-enabled capabilities that help us manage, move and keep our clients’ assets safe.”

Powered by its new system, BNY Mellon plans to use NVIDIA AI Enterprise software to support the build and deployment of AI applications and manage AI infrastructure.

NVIDIA AI Software: A Key Component in BNY Mellon’s Toolbox

Founded by Alexander Hamilton in 1784, BNY Mellon oversees nearly $50 trillion in assets for its clients and helps companies and institutions worldwide access the money they need, support governments in funding local projects, safeguard investments for millions of individuals and more.

BNY Mellon has long been at the forefront of AI and accelerated computing in the financial services industry. Its AI Hub has more than 20 AI-enabled solutions in production. These solutions support predictive analytics, automation and anomaly detection, among other capabilities.

While the firm recognizes that AI presents opportunities to enhance its processes to reduce risk across the organization, it is also actively working to consider and manage potential risks associated with AI through its robust risk management and governance processes.

Some of the use cases supported by DGX SuperPOD include deposit forecasting, payment automation, predictive trade analytics and end-of-day cash balances.

More are coming. The company identified more than 600 opportunities in AI during a firmwide exercise last year, and dozens are already in development using such NVIDIA AI Enterprise software as NVIDIA NeMo, NVIDIA Triton Inference Server and NVIDIA Base Command.

Triton Inference Server is inference-serving software that streamlines AI inferencing or puts trained AI models to work.

Base Command powers the DGX SuperPOD, delivering the best of NVIDIA software that enables businesses and their data scientists to accelerate AI development.

NeMo is an end-to-end platform for developing custom generative AI, anywhere. It includes tools for training, and retrieval-augmented generation, guardrailing and toolkits, data curation tools, and pretrained models, offering enterprises an easy, cost-effective, and fast way to adopt generative AI.

Fueling Innovation Through Top Talent

With the new DGX SuperPOD, these tools will enable BNY Mellon to streamline and accelerate innovation within the firm and across the global financial system.

Hundreds of data scientists, solutions architects and risk, control and compliance professionals have been using the NVIDIA DGX platform, which delivers the world’s leading solutions for enterprise AI development at scale, for several years.

By leveraging their new NVIDIA DGX SuperPOD will help the company rapidly expand its on-premises AI infrastructure.

The new system also underscores the company’s commitment to adopting new technologies and attracting top talent across the world to help drive its innovation agenda forward.

The NVIDIA Isaac robotics platform is tapping into the latest generative AI and advanced simulation technologies to accelerate AI-enabled robotics.

At GTC today, NVIDIA announced Isaac Manipulator and Isaac Perceptor — a collection of foundation models, robotics tools and GPU-accelerated libraries.

On stage before a crowd of 10,000-plus, NVIDIA founder and CEO Jensen Huang demonstrated Project GR00T, which stands for Generalist Robot 00 Technology, a general-purpose foundation model for humanoid robot learning. Project GR00T leverages  various tools from the NVIDIA Isaac robotics platform to create AI for humanoid robots.

“Building foundation models for general humanoid robots is one of the most exciting problems to solve in AI today,” said Huang. “The enabling technologies are coming together for leading roboticists around the world to take giant leaps toward artificial general robotics.”

NVIDIA also announced a new computer for humanoid robots based on the NVIDIA Thor system-on-a-chip, and new tools for the NVIDIA Isaac robotics platform, including Isaac Lab for robot learning and NVIDIA OSMO for hybrid-cloud workflow orchestration, which are instrumental in the development of Project GR00T and foundation models for robots.

Introducing Isaac Manipulator for Robotic Arms

NVIDIA Isaac Manipulator offers a collection of state-of-the-art motion generation and modular AI capabilities for robotic arms, with a robust collection of foundation models and GPU-accelerated libraries.

Robotics developers can use combinations of software components customized for specific tasks to perceive and interact with surroundings, enabling the building of scalable and repeatable workflows for dynamic manipulation tasks by accelerating AI model training and task programming.

“Incorporating new tools for foundation model generation into the Isaac platform accelerates the development of smarter, more flexible robots that can be generalized to do many tasks,” said Deepu Talla, vice president of robotics and edge computing at NVIDIA.

Leading robotics companies Yaskawa, Solomon, PickNik Robotics, READY Robotics, Franka Robotics, and Universal Robots, a Teradyne company, are partnering with NVIDIA to bring Isaac Manipulator to their customers.

“By bringing NVIDIA AI tools and capabilities to Yaskawa’s automation solutions, we’re pushing the boundaries of where robots can be deployed across industries,“ said Masahiro Ogawa, President,  Yaskawa. “This will significantly influence various industries.”

NVIDIA is introducing foundation models to augment existing robot manipulation systems. These will help develop robots to sense, adapt and reprogram for varied environments and applications in smart manufacturing, handling pick-and-place tasks, machine tending and assembly with the following:

• FoundationPose is a pioneering foundation model for 6D pose estimation and tracking of previously unseen objects.
• cuMotion taps into the parallel processing of NVIDIA GPUs for solving robot motion planning problems at industrial scale by running many trajectory optimizations at the same time to provide the best solution.
• FoundationGrasp is a transformer based model that can make dense grasp predictions for unknown 3D objects.
• SyntheticaDETR is an object detection model for indoor environments that allows faster detection, rendering and training with new objects.

Introducing Isaac Perceptor for Autonomous Mobile Robots Visual AI

Manufacturing and fulfillment operations are adopting autonomous mobile robots (AMRs) to improve efficiency and worker safety as well as to reduce error rates and costs.

Isaac Perceptor provides multi-camera, 360-degree vision capabilities, offering early industry partners  such as ArcBest, BYD and KION Group advanced visual AI for their AMR installations that assist in material handling operations.

The NVIDIA Nova Orin DevKit — created in collaboration with Segway Robotics and Leopard Imaging — allows companies to quickly develop, evaluate and deploy Isaac Perceptor.

“ArcBest is collaborating with NVIDIA to bring leading-edge machine vision technology into the logistics space,” said Michael Newcity, chief innovation officer of ArcBest and president of ArcBest Technologies. “Using the Isaac Perceptor platform in our Vaux Smart Autonomy AMR forklifts and reach trucks enables better perception, semantic-aware navigation and 3D mapping for obstacle detection in material handling processes across warehouses, distribution centers and manufacturing facilities.”

Project GR00T for Humanoid Robotics Development Takes a Bow

Demonstrated at GTC, GR00T-powered humanoid robots can take multimodal instructions — text, video and demonstrations — as well as their previous interactions to produce the desired action for the robot. GR00T was shown on four humanoid robots from different companies, including Agility Robotics, Apptronik, Fourier Intelligence and Unitree Robotics.

Humanoid robots are complex systems that require heterogeneous computing to meet the needs of high frequency low level controls, sensor fusion and perception, task planning and human-robot interaction. NVIDIA unveiled a new Jetson Thor-based computer for humanoid robots, built on the NVIDIA Thor SoC.

Jetson Thor includes a next-generation GPU based on the NVIDIA Blackwell Architecture with a transformer engine delivering 800 teraflops of 8-bit floating point AI performance to run multimodal generative AI models like GR00T. With an integrated functional safety processor, a high-performance CPU cluster and 100GB of ethernet bandwidth, it significantly simplifies design and integration efforts.

Project GR00T uses Isaac tools that are available to robotics developers for building and testing foundation models. These include Isaac Lab, a new lightweight simulation app built in Isaac Sim to train this humanoid robot model at scale, and OSMO, a cloud workflow orchestration platform for managing the training and simulation workloads.

Accelerating Robot Learning With Isaac Lab

Robots that require advanced locomotion skills, whether with walking or grasping, need to use deep reinforcement learning in a simulated environment and be trained repeatedly in a virtual environment to learn skills. However, this utility becomes more useful when the model transfers to the real robot deployment, which has been demonstrated with Project GR00T.

As the successor to Isaac Gym, Isaac Lab benefits from NVIDIA Omniverse technologies for physics-informed, photorealistic, perception-based reinforcement learning tasks. Isaac Lab is an open-source, performance-optimized application for robot learning built on the Isaac Sim platform. It incorporates reinforcement learning APIs and a developer-friendly tasking framework.

Enabling Cloud-Native Robotics Workflow Scheduling With NVIDIA OSMO 

NVIDIA OSMO scales workloads across distributed environments. For robotics workloads with complex multi-stage and multi-container workflows, the platform provides a location-agnostic deployment option and dataset management and traceability features for deployed models.

“Boston Dynamics employs a range of machine learning, reinforcement learning and AI technologies to power our robots,” said Pat Marion, machine learning and perception lead at Boston Dynamics. “To effectively manage the large training workloads, we’re using NVIDIA OSMO, an infrastructure solution that lets our machine learning engineers streamline their workflows and dedicate their expertise to tackling the hard robotics problems.”

OSMO supports GR00T, for example, by concurrently running models on NVIDIA DGX for training and NVIDIA OVX servers for live reinforcement learning in simulation. This workload involves generating and training models iteratively in a loop. OSMO’s ability to manage and schedule workloads across distributed environments allows for the seamless coordination of DGX and OVX systems, enabling efficient and iterative model development. Once the model is ready for testing and validation, OSMO can uniquely orchestrate software-in-the-loop workflows on OVX (x86-64) as well as hardware-in-the-loop workflows with NVIDIA Jetson (aarch64) compute resources.

Supporting the ROS Ecosystem of Developers

NVIDIA joined the Open Source Robotics Alliance (OSRA) as a founding member and platinum sponsor. OSRA is a new initiative by Open Source Robotics Foundation to foster collaboration, innovation and technical guidance in the robotics community by supporting several open-source robotics projects, including the Robot Operating System (ROS).

“The increasing capability of autonomous robots is driving a rise in demand for more powerful but still energy-efficient onboard computing,” said Vanessa Yamzon Orsi, CEO of Open Robotics. “The ROS community is experiencing this demand firsthand, and our users are increasingly taking advantage of advanced accelerated computing hardware from industry leaders such as NVIDIA.”

NVIDIA Isaac Perceptor with Nova Orin evaluation kit, Isaac Manipulator, Isaac Lab and OSMO will be made available to customers and partners in the second quarter of this year. Learn more about Project GR00T.