{"id":1175,"date":"2021-11-10T08:34:58","date_gmt":"2021-11-10T08:34:58","guid":{"rendered":"https:\/\/salarydistribution.com\/machine-learning\/2021\/11\/10\/gtc-wrap-up-nvidia-ceo-outlines-vision-for-accelerated-computing-data-center-architecture-ai-robotics-omniverse-avatars-and-digital-twins-in-keynote\/"},"modified":"2021-11-10T08:34:58","modified_gmt":"2021-11-10T08:34:58","slug":"gtc-wrap-up-nvidia-ceo-outlines-vision-for-accelerated-computing-data-center-architecture-ai-robotics-omniverse-avatars-and-digital-twins-in-keynote","status":"publish","type":"post","link":"https:\/\/salarydistribution.com\/machine-learning\/2021\/11\/10\/gtc-wrap-up-nvidia-ceo-outlines-vision-for-accelerated-computing-data-center-architecture-ai-robotics-omniverse-avatars-and-digital-twins-in-keynote\/","title":{"rendered":"GTC Wrap-Up: NVIDIA CEO Outlines Vision for Accelerated Computing, Data Center Architecture, AI, Robotics, Omniverse Avatars and Digital Twins in Keynote"},"content":{"rendered":"
\n

Bringing simulation of real and virtual worlds to everything from self-driving vehicles to avatars to robotics to modeling the planet\u2019s climate, NVIDIA founder and CEO Jensen Huang<\/a> Tuesday introduced technologies to transform multitrillion-dollar industries.<\/p>\n

Huang delivered a keynote at the company\u2019s virtual GTC<\/a> gathering where he unveiled NVIDIA Omniverse Avatar<\/a> and NVIDIA Omniverse Replicator<\/a>, among a host of announcements, demos and far-reaching initiatives.<\/p>\n

Huang showed how NVIDIA Omniverse<\/a> \u2014 the company\u2019s virtual world simulation and collaboration platform for 3D workflows \u2014 brings NVIDIA\u2019s technologies together.<\/p>\n

And he showed a demonstration of Project Tokkio for customer support and Project Maxine for video conferencing with Omniverse Avatar.<\/p>\n

<\/p>\n

\"\"<\/p>\n

<\/a>

NVIDIA CEO Jensen Huang showed how Project Maxine for Omniverse Avatar connects computer vision, Riva speech AI, and avatar animation and graphics into a real-time conversational AI robot \u2014 the Toy Jensen Omniverse Avatar.<\/figcaption><\/figure>\n

\u201cA constant theme you\u2019ll see \u2014 how Omniverse is used to simulate digital twins of warehouses, plants and factories, of physical and biological systems, the 5G edge, robots, self-driving cars, and even avatars,\u201d Huang said.<\/p>\n

Huang ended by announcing NVIDIA will build a digital twin, called E-2, or Earth Two, to simulate and predict climate change.<\/p>\n

\u2018Full Stack, Data Center Scale, and Open Platform\u2019<\/h2>\n

Accelerated computing<\/a> launched modern AI, and the waves it started are coming to science and the world\u2019s industries, Huang said.<\/p>\n

That starts with three chips, the GPU, CPU<\/a> and DPU<\/a>, and systems \u2014 DGX<\/a>, HGX<\/a>, EGX<\/a>, RTX<\/a> and AGX<\/a> \u2014 spanning from cloud to the edge, he said.<\/p>\n

NVIDIA has created 150 acceleration libraries for 3 million developers, from graphics and AI to sciences and robotics.<\/p>\n

And NVIDIA is introducing 65 new and updated SDKs<\/a> at GTC.<\/p>\n

\u201cNVIDIA accelerated computing is a full-stack, data-center-scale and open platform,\u201d Huang said.<\/p>\n

Huang introduced NVIDIA Quantum-2<\/a>, \u201cthe most advanced networking platform ever built,\u201d and with the BlueField-3 DPU, welcomes cloud-native supercomputing.<\/p>\n

Quantum-2 offers the extreme performance, broad accessibility and strong security needed by cloud computing providers and supercomputing centers, he said.<\/p>\n

Cybersecurity is a top issue for companies and nations, and Huang announced a three-pillar zero-trust framework to tackle the challenge.<\/p>\n<\/p>\n

BlueField DPUs<\/a> isolate applications from infrastructure. NVIDIA DOCA 1.2 \u2014 the latest BlueField SDK \u2014 enables next-generation distributed firewalls. NVIDIA Morpheus<\/a>, assuming the interloper is already inside, uses the \u201csuperpowers of accelerated computing and deep learning to detect intruder activities,\u201d Huang said.<\/p>\n

Omniverse Avatar and Omniverse Replicator<\/h2>\n

With Omniverse, \u201cwe now have the technology to create new 3D worlds or model our physical world,\u201d Huang said.-<\/p>\n

To help developers create interactive characters with Omniverse that can see, speak, converse on a wide range of subjects and understand naturally spoken intent, NVIDIA announced Omniverse Avatar.<\/p>\n

Huang showed how Project Maxine for Omniverse Avatar connects computer vision, Riva speech AI, and avatar animation and graphics into a real-time conversational AI robot \u2014 the Toy Jensen Omniverse Avatar.<\/p>\n<\/p>\n

He also showed a demo of Project Tokkio, a customer-service avatar in a restaurant kiosk, able to see, converse with and understand two customers.<\/p>\n

Huang additionally showed Project Maxine\u2019s ability to add state-of-the-art video and audio features to virtual collaboration and content creation applications.<\/p>\n

A demo showed a woman speaking English on a video call in a noisy cafe, but she can be heard clearly without background noise. As she speaks, her words are transcribed and translated in real time into French, German and Spanish. And, thanks to Omniverse, they\u2019re spoken by an avatar able to engage in conversation with her same voice and intonation.<\/p>\n<\/p>\n

To help developers to create the huge amounts of data needed to train AI, NVIDIA announced Omniverse Replicator<\/a>,\u00a0 a synthetic-data-generation engine for training deep neural networks.<\/p>\n

NVIDIA has developed two replicators: one for general robotics, Omniverse Replicator for Isaac Sim, and another for autonomous vehicles, Omniverse Replicator for DRIVE Sim<\/a>.<\/p>\n<\/p>\n

Since its launch late last year, Omniverse has been downloaded 70,000 times by designers at 500 companies.<\/p>\n

Omniverse Enterprise is now available starting at $9,000 a year.<\/p>\n<\/p>\n

AI Models and Systems<\/h2>\n

Huang introduced Nemo Megatron to train large language models<\/a>. Such models \u201cwill be the biggest mainstream HPC application ever,\u201d he said.<\/p>\n

Graphs \u2014 a key data structure in modern data science \u2014 can now be projected into deep-neural network frameworks with Deep Graph Library, or DGL, a new Python package.<\/p>\n

NVIDIA Modulus<\/a>, introduced Tuesday, builds and trains physics-informed machine learning models that can learn and obey the laws of physics.<\/p>\n

An upgrade to Triton<\/a>, an inference server for all workloads, now inferences forest models and does multi-GPU multi-Node inference for large language models.<\/p>\n

And Huang introduced three new libraries.<\/p>\n