search

Nvidia introduces set of new products to tackle growing AI demands

US firm introduces a first central processing unit (CPU) for data centers, with 10x performance of the fastest servers; and next-gen of Israel-developed data processing unit

Shoshanna Solomon is The Times of Israel's Startups and Business reporter

Nvidia's founder and CEO Jensen Huang speaking at the firm's GTC developers conference in Tel Aviv on October 18, 2017. (courtesy)
Nvidia's founder and CEO Jensen Huang speaking at the firm's GTC developers conference in Tel Aviv on October 18, 2017. (courtesy)

US gaming and computer graphics giant Nvidia Corp., which completed the acquisition of Israel’s Mellanox Technologies, Ltd. last year for a massive $7 billion, announced on Monday a series of new products which it hopes will give it an edge over competitors.

Mellanox teams in Israel, now a division within the US firm called Nvidia Networking, were part of the development of a number of the products, Gilad Shainer, senior vice president of High Performance Computing at the firm said in a press briefing with reporters. One of the key products the Israeli team has worked on is a new data processing unit (DPU) called BlueField-3.

Nvidia CEO Jensen Huang presented the technologies at the online technology conference GTC21 event held by the firm on Monday.

The company, which is focusing on developing processing units to speed up developments in artificial intelligence, introduced on Monday its first central processing unit (CPU) for data centers, called the Grace, which will deliver 10 times the performance of today’s fastest servers.

Nvidia’s Omniverse platform allows teams designing 3D products to work together from around the world (Courtesy).

The firm also unveiled a next-generation AI-enabled processor for autonomous vehicles, the Nvidia Drive Atlan, which will deliver more than 1,000 trillion operations per second, and targets automakers’ 2025 models.

A new framework created by the firm, Jarvis, will provide developers with pre-trained deep learning models and software tools for interactive conversational AI services; while a new generation of its data processing unit (DPU), BlueField-3, developed by the Israel-based team will deliver software-defined networking, storage, and cybersecurity acceleration capabilities for data centers.

The US firm also announced a new platform, Omniverse, that allows teams designing 3D products to work together from different locations around the world. And the Nvidia Morpheus, that brings AI automation capabilities to the cybersecurity industry.

The result of more than 10,000 engineering years of work, the Nvidia “Grace” CPU is designed to address the computing requirements for the world’s most advanced artificial intelligence applications – including natural language processing, recommender systems, and AI supercomputing, the company said.

AI processes analyze enormous datasets and require both ultra-fast computing performance and massive memory. Grace combines energy-efficient Arm CPU cores with an innovative low-power memory subsystem to deliver high performance, with an energy-efficient design, the company said.

Nvidia Jarvis will enable a new wave of language-based applications, improving interactions with humans and machines. (courtesy)

The processor is named for Grace Hopper, a US computer programming pioneer.

The Swiss National Supercomputing Centre (CSCS) and the US Department of Energy’s Los Alamos National Laboratory are the first to announce plans to build Grace-powered supercomputers for their national scientific research efforts, the company said.

“Leading-edge AI and data science are pushing today’s computer architecture beyond its limits, processing unthinkable amounts of data,” said Nvidia’s CEO Huang.

Availability of the Grace processor is expected in the beginning of 2023, the company said.

Nvidia Jarvis will enable a new wave of language-based applications, improving interactions with humans and machines, the company said. It opens the door to the creation of such services as digital nurses to help monitor patients around the clock, relieving overloaded medical staff; online assistants to understand what consumers are looking for and recommend the best products; and real-time translations to improve cross-border workplace collaboration and enable viewers to enjoy live content in their own language.

Jarvis has been built using models trained for several million graphics processing unit (GPU) hours on over 1 billion pages of text, 60,000 hours of speech data, and in different languages, accents, environments, and lingos to achieve high levels of accuracy, the company said.

read more:
comments
Never miss breaking news on Israel
Get notifications to stay updated
You're subscribed