Meet the $10,000 Nvidia chip powering the race for AI

The Nvidia A100 chip powers most of the applications several tech giants use to integrate cutting-edge AI into their search engines. It is a $10,000 chip that is the most critical tool in the AI sector now.

The Powerhouse of AI

Nathan Benaich, an AI reporter and investor, says many AI professionals rely on this mighty chip to develop supercomputers. New Street Research also confirms that at least 95% of graphic processors for machine learning use the A100. New Power tools like ChatGPT, Bing AI, and Stable Diffusion rely on the A100.

The chip performs simple calculations for training neural network systems. Initially, A100 acted as a graphics processor unit created for 3D rendering in video games. Programmers reconfigured it for machine learning tasks in data centers.

Demand for A100

Large artificial intelligence applications require hundreds of graphic processors for inference and learning how to generate texts, translate languages, or identify objects on the web.
Based on this alone, tech giants need thousands of Nvidia’s A100 to operate smoothly and improve their AI models. Emad Mostaque, Stability AI CEO, tweeted in January that the company uses the number of A100s it has to gauge its progress. He explained that in 2022, the company had less than 35 A100s, and today, they have direct and remote access to about 5,400 GPUs.

Nvidia results show its growing lead in AI chip race https://t.co/7AT6frZvsx pic.twitter.com/3USB9nm3HO

— Reuters (@Reuters) February 23, 2023

Nvidia Benefits Immensely

According to their fiscal earnings report, Nvidia’s AI chip business recorded over $3.6 billion last quarter. Investors continue to push the stock up this year.
So far, Nvidia stocks are up 65% in the past eight weeks. Overtaking major S&P 500 and semiconductors stocks.

In a discussion with analysts about the company’s strategy, CEO Jensen Huang referred to the increase in activity surrounding the AI infrastructure Nvidia has constructed and the usage of their Hopper and Ampere CPUs for sizable language models. Huang highlighted recently launched Hopper generation and Ampere generation chips, which feature the H100 and A100 models.

He certainly gave the impression that Nvidia is very focused on AI and regards it as a crucial component of its corporate strategy.

Featured image from Yahoo.com