Jornal Do Brasilia - Nvidia, the world's newest, AI-amped tech giant

- Decades-old upstart -

Nvidia is not an out-of-the-blue startup.

Founded in 1993, Nvidia designs chips that are used in the fastest developing sectors of the tech business: gaming, video-editing, self-driving cars and, now, artificial intelligence. Its technology was also in the mix for the crypto boom.

"We had this idea that computer graphics was going to be the driving force of technology and [its] fuel would be video games," co-founder and CEO Jensen Huang said in 2018.

Based in California, Nvidia doesn't actually make its own chips, but rather designs them and then outsources the manufacturing to other companies, most notably Taiwan Semiconductor Manufacturing Company.

Its chips, known as graphics processing units (GPU's), were used to create the effects in "Avatar" and other blockbuster films. But Nvidia turned into a behemoth when its wares proved to be adaptable to other industries that need huge computing power.

It also builds the systems and software that run its products, modelling its business plan on Apple, which uses must-have hardware to rope in consumers to other services.

- Right product, right time -

Nvidia's bread and butter has been the GPU and for the first decades of its existence, the company was laser-focused on delivering the best possible graphics for video games and movies.

There's only one final judge and "it's the human eye," Chris Malachowsky, another Nvidia co-founder, said in 2012.

But soon, the chip was also seen as effective for other uses, including mining crypto currencies, processing massive amounts of data, and machine learning, the heavy computing process behind the AI revolution.

As the use cases expanded, and ChatGPT conquered the world, the company only grew stronger and it now holds an 82 percent market share for standalone GPUs.

In 2022, Nvidia released the H100, one of the most powerful processors it has ever built, costing about $40,000 each, which it said was the first chip designed specifically for generative AI.

The H100, which holds 80 billion transistors, is seeing exploding demand from the cloud giants that power the AI arms race, such as Microsoft, Amazon and Google and any other company that can afford to join battle.

Elon Musk last week said that GPUs "are considerably harder to get than drugs" at the moment and the dependence is a rainmaker for Nvidia.

Nvidia announced this month that its sales for the three months ending in July would be an eye-watering $11 billion.

- Leather jacket -

What Steve Jobs did for the turtle neck, Nvidia's hard-charging Huang is trying to do for the leather jacket.

At product launches, the 60-year-old Taiwanese-American immigrant sports a leather motorcycle jacket and is known to make video gags sporting the coat to plug new releases.

Born in Taiwan, his parents sent him to a strict boarding school in Kentucky in the 1970s where Huang said he and his brothers learned to survive in a tough environment.

Huang later earned engineering degrees at Oregon State University and Stanford University.

Last week Huang had a hero's homecoming in Taiwan where he said the world was at "the tipping point of a new computer era."

- Meme stock -

For a while, Nvidia was an unsung hero of the tech industry and even became a meme stock, pumped up by day traders on social media, when it was still largely overlooked by the bigwigs on Wall Street.

S. Soares--JDB