If there’s one company that truly rode the AI wave to the top, it’s Nvidia. Today, the company is at the center of this technological revolution, with its chips powering everything from chatbots to massive data centers. If your company needs to train new AI systems, you need Nvidia chips. If you need to deliver AI models to consumers, you need an Nvidia chip. In a sense.
But how did Nvidia get here? Nvidia CEO Jensen Huang shares the story. Before 2016, Nvidia looked very different. It was certainly well known, but mainly within the gaming community. But Nvidia has bigger plans for the future, and its story actually started in a small room.
In a podcast with Joe Rogan, CEO Jensen Huang revealed that the company’s turning point was when it started betting big on deep learning technology long before other companies in the industry took it on. “We knew we needed to put our all into this and help create this industry because we were going to solve a lot of interesting problems, and that was in 2012,” he said.
Huang went on to explain that Nvidia built the first true AI machine, DGX-1, based on this idea. It was designed specifically for deep learning and had a hefty price tag of around $300,000. “DGX-1 cost $300,000. It cost Nvidia billions of dollars to make the first one. And instead of two chips, SLI, they connected eight chips with a technology called NVLink, which is basically a supercharged version of SLI,” Huang revealed.
As gamers will remember, SLI was a technology that allowed computer gamers to use two graphics cards at the same time by connecting them using ribbon cables.
Nvidia’s machines were powerful and ahead of their time, but there was just one problem. No one wanted it. Huang said that when NVIDIA announced DGX-1, there were no buyers. Most companies didn’t understand what it was, let alone why they needed it.
“When I announced this thing, no one in the world wanted it. I didn’t have a purchase order. There wasn’t one. No one wanted to buy it. No one wanted to be a part of it. The audience was completely silent. They didn’t know what I was saying,” he said.
first customer
But everything changed when Elon Musk showed interest. Huang recalled that Musk contacted him and told him there was a company that could use such a machine. There was just one problem. That means it was non-profit. That company later became OpenAI.
Despite the high cost and uncertainty, Nvidia decided to take a leap of faith. In 2016, Huang personally delivered DGX-1 to Musk in San Francisco. But he experienced a surprise. The customer for this advanced supercomputer was not a state-of-the-art laboratory, but a small group of researchers working in a small room. Among them was Ilya Satskeva, who would later play a key role in the formation of modern AI.
“I walked up to the second floor, where it felt like I was in a smaller room than you are here, and it turned out to be OpenAI 2016, just a bunch of people sitting in the same room,” Huang said.
But 10 years later, one machine in that tiny room changed everything. This laid the foundation for ChatGPT and pushed generative AI into the mainstream.
Now, Nvidia is more than just a chipmaker. It is the owner of the engine behind every major LLM on the planet. Well, at least almost all of the AI models don’t exist in China or come from Google. Google uses its own Tensor Processing Units (TPUs). With the explosive development of AI, Nvidia has become the world’s largest company by market capitalization.
– end
