{"id":2965,"date":"2023-05-01T14:11:23","date_gmt":"2023-05-01T14:11:23","guid":{"rendered":"https:\/\/salarydistribution.com\/machine-learning\/2023\/05\/01\/now-shipping-dgx-h100-systems-bring-advanced-ai-capabilities-to-industries-worldwide\/"},"modified":"2023-05-01T14:11:23","modified_gmt":"2023-05-01T14:11:23","slug":"now-shipping-dgx-h100-systems-bring-advanced-ai-capabilities-to-industries-worldwide","status":"publish","type":"post","link":"https:\/\/salarydistribution.com\/machine-learning\/2023\/05\/01\/now-shipping-dgx-h100-systems-bring-advanced-ai-capabilities-to-industries-worldwide\/","title":{"rendered":"Now Shipping: DGX H100 Systems Bring Advanced AI Capabilities to Industries Worldwide"},"content":{"rendered":"
\n

Customers from Japan to Ecuador and Sweden are using NVIDIA DGX H100<\/a> systems like AI factories to manufacture intelligence.<\/p>\n

They\u2019re creating services that offer AI-driven insights in finance, healthcare, law, IT and telecom \u2014 and working to transform their industries in the process.<\/p>\n

Among the dozens of use cases, one aims to predict how factory equipment will age, so tomorrow\u2019s plants can be more efficient.<\/p>\n

Called Green Physics AI<\/a>, it adds information like an object\u2019s CO2 footprint, age and energy consumption to SORDI.ai<\/a>, which claims to be the largest synthetic dataset in manufacturing.<\/p>\n

\"Green<\/a>
Green Physics AI lets users model how objects age.<\/figcaption><\/figure>\n

The dataset lets manufacturers develop powerful AI models and create digital twins that optimize the efficiency of factories and warehouses.\u00a0 With Green Physics AI, they also can optimize energy and CO2 savings for the factory\u2019s products and the components that go into them.<\/p>\n

Meet Your Smart Valet<\/b><\/h2>\n

Imagine a robot that could watch you wash dishes or change the oil in your car, then do it for you.<\/p>\n

Boston Dynamics AI Institute<\/a> (The AI Institute), a research organization which traces its roots to Boston Dynamics, the well-known pioneer in robotics, will use a DGX H100 to pursue that vision. Researchers imagine dexterous mobile robots helping people in factories, warehouses, disaster sites and eventually homes.<\/p>\n

\u201cOne thing I\u2019ve dreamed about since I was in grad school is a robot valet who can follow me and do useful tasks \u2014 everyone should have one,\u201d said Al Rizzi, CTO of The AI Institute.<\/p>\n

That will require breakthroughs in AI and robotics, something Rizzi has seen firsthand. As chief scientist at Boston Dynamics, he helped create robots like Spot, a quadruped that can navigate stairs and even open doors for itself.<\/p>\n

Initially, the DGX H100 will tackle tasks in reinforcement learning<\/a>, a key technique in robotics. Later, it will run AI inference jobs while connected directly to prototype bots in the lab.<\/p>\n

\u201cIt\u2019s an extremely high-performance computer in a relatively compact footprint, so it provides an easy way for us to develop and deploy AI models,\u201d said Rizzi.<\/p>\n

Born to Run Gen AI<\/b><\/h2>\n

You don\u2019t have to be a world-class research outfit or Fortune 500 company to use a DGX H100. Startups are unboxing some of the first systems to ride the wave of generative AI<\/a>.<\/p>\n

For example, Scissero<\/a>, with offices in London and New York, employs a GPT-powered chatbot to make legal processes more efficient. Its Scissero GPT can draft legal documents, generate reports and conduct legal research.<\/p>\n

In Germany, DeepL<\/a> will use several DGX H100 systems to expand services like translation between dozens of languages it provides for customers, including Nikkei, Japan\u2019s largest publishing company. DeepL recently released an AI writing assistant called DeepL Write.<\/p>\n

Here\u2019s to Your Health<\/b><\/h2>\n

Many of the DGX H100 systems will advance healthcare and improve patient outcomes.<\/p>\n

In Tokyo, DGX H100s will run simulations and AI to speed the drug discovery process as part of the Tokyo-1 supercomputer<\/a>. Xeureka<\/a> \u2014 a startup launched in November 2021 by Mitsui & Co. Ltd., one of Japan\u2019s largest conglomerates \u2014\u00a0 will manage the system.<\/p>\n

Separately, hospitals and academic healthcare organizations in Germany, Israel and the U.S. will be among the first users of DGX H100 systems.<\/p>\n

Lighting Up Around the Globe<\/b><\/h2>\n

Universities from Singapore to Sweden are plugging in DGX H100 systems for research across a range of fields.<\/p>\n

A DGX H100 will train large language models<\/a> for Johns Hopkins University Applied Physics Laboratory<\/a>. The KTH Royal Institute of Sweden will use one to expand its supercomputing capabilities.<\/p>\n

Among other use cases, Japan\u2019s CyberAgent, an internet services company, is creating smart digital ads and celebrity avatars. Telconet, a leading telecommunications provider in Ecuador, is building intelligent video analytics for safe cities and language services to support customers across Spanish dialects.<\/p>\n

An Engine of AI Innovation<\/b><\/h2>\n

Each NVIDIA H100 Tensor Core GPU<\/a> in a DGX H100 system provides on average about 6x more performance<\/a> than prior GPUs. A DGX H100 packs eight of them, each with a Transformer Engine<\/a> designed to accelerate generative AI models.<\/p>\n

The eight H100 GPUs connect over NVIDIA NVLink<\/a> to create one giant GPU. Scaling doesn\u2019t stop there: organizations can connect hundreds of DGX H100 nodes into an AI supercomputer using the 400 Gbps ultra-low latency NVIDIA Quantum InfiniBand<\/a>, twice the speed of prior networks.<\/p>\n

Fueled by a Full Software Stack<\/h2>\n

DGX H100 systems run on NVIDIA Base Command<\/a>, a suite for accelerating compute, storage, and network infrastructure and optimizing AI workloads.<\/p>\n

They also include NVIDIA AI Enterprise<\/a>, software to accelerate data science pipelines and streamline development and deployment of generative AI, computer vision and more.<\/p>\n

The DGX platform offers both high performance and efficiency. DGX H100 delivers a 2x improvement in kilowatts per petaflop over the DGX A100 generation.<\/p>\n

NVIDIA DGX H100 systems, DGX PODs and DGX SuperPODs are available from NVIDIA\u2019s global partners<\/a>.<\/p>\n

Manuvir Das, NVIDIA\u2019s vice president of enterprise computing, announced DGX H100 systems are shipping in a talk at MIT Technology Review\u2019s Future Compute<\/a> event today. A link to his talk will be available here soon.<\/p>\n<\/div>\n","protected":false},"excerpt":{"rendered":"

https:\/\/blogs.nvidia.com\/blog\/2023\/05\/01\/dgx-h100-systems-shipping\/<\/p>\n","protected":false},"author":0,"featured_media":2966,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":[],"categories":[3],"tags":[],"_links":{"self":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/posts\/2965"}],"collection":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/types\/post"}],"replies":[{"embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/comments?post=2965"}],"version-history":[{"count":0,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/posts\/2965\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/media\/2966"}],"wp:attachment":[{"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/media?parent=2965"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/categories?post=2965"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/salarydistribution.com\/machine-learning\/wp-json\/wp\/v2\/tags?post=2965"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}