T

he business world is increasingly banking on artificial intelligence to be the next big thing, and has found itself turning to one maker of computer chips in particular — Nvidia — to power the revolution. Since 1993, the Santa Clara, California-based company has been designing programmable chips that help run an array of consumer-facing applications. While Intel and Advanced Micro Devices had dominated the U.S. chip sector for decades, Nvidia’s entry signaled the advent of sophisticated graphics processing units (GPUs), which were better able to render images. That capability became ever more important as high-quality video increasingly dominated the tech and media landscape. At first, Nvidia was most associated with providing GPU processors for video game consoles like the Microsoft Xbox and Sony PlayStation. The general growth of Silicon Valley during the 2010s prompted Nvidia to diversify and improve its fortunes. For example, in 2014, Nvidia and Google announced a partnership to use Nvidia chips in Google Chromebooks. Auto companies also began turning to Nvidia chips for use in driver-assistance software that depends on GPUs to process image information from sensors. Nvidia hardware is also found in all Tesla vehicles. During the pandemic, the shift to remote work and subsequent demand for data centers that could enable cloud-based computing — plus even more interest in video games while everyone was stuck indoors — accelerated Nvidia’s revenues even further. Still, the company’s $22 billion in annual revenues in 2022 remained dwarfed by rival Intel’s $63 billion that year. It was not until the last year and a half or so that Nvidia’s outlook went truly stratospheric with the onset of the artificial intelligence revolution. The chips that Nvidia specializes in, known as “discrete” GPUs, specialize in so-called parallel computing. This is a type of computing in which multiple processes are carried out simultaneously. By contrast, CPUs carry out programs serially, or sequentially. The type of computing GPUs handle is thus more efficient and more valuable than CPUs. Programmers have learned that Nvidia’s GPUs are much better suited for programming AI software. Bloomberg News has called Nvidia’s chips the “workhorse for training AI models,” and PNC Financial Services Group analyst Amanda Agati described Nvidia’s lead in the category last fall, based on its valuation, as a “quasi monopoly.” So Nvidia owes some of its success to having pioneered the technology that AI applications now depend upon. Today, virtually every major tech company, including Amazon, Google, Meta, Microsoft and Oracle, has made use of Nvidia chips. On Thursday, Nvidia reported earnings and revenues that significantly surpassed market expectations, helping lift the company’s overall value past $2 trillion, trailing only Microsoft and Apple among the largest U.S. firms. “Accelerated computing and generative AI have hit the tipping point,” Nvidia CEO and co-founder Jensen Huang said in the earnings release. “Demand is surging worldwide across companies, industries and nations.” About $1.7 trillion of the market capitalization growth has come in the past 16 months alone, according to Goldman Sachs. Huang acknowledged in an interview with CNBC last year that a combination of luck and skill has led to the company’s success.