Nvidia Announces New Data Center Influence With Turing GPUs



[ad_1]

After giving its long-time fans the first look at its new graphics processing units, Nvidia Corp. will now use its new Turing technology to bolster its artificial intelligence efforts.

During a speech in Japan on Thursday, general manager Jensen Huang announced a new AI inference platform in the data center that uses the Turing GPUs announced for the first time last month. Nvidia

NVDA, -1.69%

also announced new Turing-based systems for machines in addition to autonomous cars, with an eye on manufacturing robots and healthcare machines that process a lot of data.

While Nvidia has become a leader in machine learning in recent years, it has mostly used half of the AI ​​training, researchers using the processing power offered by Nvidia GPUs to power the machines. On the inference side, Nvidia has announced a new Tesla T4 chip for servers and TensorRT software designed to strengthen inference capabilities in data centers.

"Our AI inference market has exploded," said Ian Buck, who heads Nvidia's data center business, at a press conference on Wednesday afternoon.

Nvidia's ability to develop a strong focus on inference and training is important "because training evolves with development, while inference evolves with subsequent implementation," writes this year Morgan Stanley analysts. Be part of the inference to obtain long-term sustainable rewards from his work in training.

Morgan Stanley revised Nvidia's overweight position back in April, mainly because analysts thought Nvidia was making more progress in terms of AI inference, which generally implies higher volumes but higher returns. lower margins. Nvidia said on Wednesday that the inference market would amount to $ 20 billion in the next five years.

"We now think that hardware and software developments have positioned Nvidia to capture more of the inference, key to the long-term growth rate," analysts wrote in April.

Nvidia has stated that Alphabet Inc.

GOOGL, -1.55%

GOOG, -1.23%

Google had already agreed to deploy T4 chips in its data centers, and noted in a press release that several other cloud providers and major server companies – including Microsoft Corp.

MSFT, + 0.42%

, Cisco Systems Inc.

CSCO, -0.30%

, Dell Technologies Inc., International Business Machines Corp.

IBM, + 0.05%

and Hewlett Packard Enterprise Co.

HPE, -0.30%

– expressed support for the platform.

"AI is becoming more ubiquitous, and inference is a critical capability that customers need to successfully deploy their AI models, so we're excited to soon support Nvidia's Turing Tesla GPUs on the Google Cloud Platform." said in the statement.

Nvidia has also announced new high-performance Jetson AGX Xavier compute systems for stand-alone robots and other machines, while continuing to look for ways to exploit battery life beyond cars. The announcement was intended for the audience for the Japanese version of Nvidia's Gv conference and included partnerships with Japanese companies such as Yamaha Motor Corp.

7272, + 1.82%

Nvidia said it plans to use the platform on "unmanned agricultural vehicles, last-mile vehicles and marine products".

In addition, a special version of the AGX specifically designed for the healthcare industry has been announced with accompanying software, with all these products nicknamed Clara. Nvidia announced plans to focus on medical imaging and other data-related health tasks at its main Silicon Valley GTC conference in March.

Nvidia shares fell 1.7% at Wednesday's regular session, but rose 38.6% so far this year as the S & P 500 index rose 8.1%.

Get the stories of the day delivered to your inbox. Subscribe to MarketWatch's free Tech Daily newsletter. register here

[ad_2]
Source link