Semiconductor industry bets on artificial intelligence

After some difficult years for the semiconductor industry, which is recovering from long supply shortages and is starring in a battle between USA and China due to its control, this strategic sector begins to see in the artificial intelligence a horizon of stability.

The American company Nvidia dominates the market for specialized microchips known as GPUs, ideal for training artificial intelligence (AI) programs such as the popular chatbot ChatGPT.

“Technology trends are going in the direction of Nvidia”the president of the company, Ronnie Vasishta, told AFP this week during the Mobile World Congress (MWC) in Barcelona.

This helped make Nvidia the largest company in the industry – and one of the largest companies in the United States – with a value of 580,000 million dollars.

Thus, traditional rivals like Intel and Qualcomm are working equally hard to make sure they aren’t left out.

These tiny components are essential to just about everything, from smartphones, computers, and electric cars to sophisticated weapons, robotics, and other high-tech machinery.

AI already has a large presence in these fields, but the advent of chatbots has further extended its potential and the enthusiasm of the sector.

Potential

“The most exciting thing right now is the AI”, Cristiano Amon, head of rival firm Qualcomm, said at a Wall Street Journal event at MWC.

He would like phones to have microchips capable of performing the most complex AI tasks, especially since Qualcomm is a leader in the field of phone chips.

Similarly, Vasishta is also excited.

“Where and how is AI used? It will probably be easier to answer where it is not being used.” explained.

Arm, a UK-based semiconductor company, still works further down the production chain than Nvidia, making the designs used by chip vendors.

The type of microchips the company produces are very effective for training AI models in data centers, said Chris Bergey, the company’s vice president, but smartphones need chips that can act on those models.

“It’s a huge opportunity and it’s ubiquitous,” said Bergey, who believes that AI has enormous potential.

The AI ​​revolution is for him comparable to the advent of apps, which appeared about 15 years ago and rapidly changed the way we use technology.

“AI is definitely something that has a lot of interesting applications and we’re still just getting the surface of where it’s going to go.”

Attention

With semiconductors, however, nothing is simple.

The supply chain is very complex: Accenture consultancy estimates that a microchip crosses borders 70 times before ending up in a phone, camera or car.

Countries like China and the United States would like to have more control, but there is another problem added: the factories that produce the majority of semiconductors are in Taiwan, an autonomous island that China claims, which could bring Beijing and Washington into direct conflict.

Usually prudent, semiconductor executives rarely discuss these matters.

“We do not have a position on geopolitics, we comply with all US regulations that are required of us as a US company,” Vasishta explained.

Bergey, who has spent 25 years in the industry, said he had already seen chips go from being “very interesting” to “very boring”.

“Now they are interesting, perhaps too much and with too much attention”he claimed.

“It’s a dynamic issue that the industry is handling and we have to see how it plays out.”

Source: AFP